blob: 24969be8870d98813bbb7c71cef8445a6c9a9bc2 [file] [log] [blame]
Willy Tarreau2212e6a2015-10-13 14:40:55 +02001 ------------------------
2 HAProxy Management Guide
3 ------------------------
Willy Tarreau2e077f82019-11-25 20:36:16 +01004 version 2.2
Willy Tarreau2212e6a2015-10-13 14:40:55 +02005
6
7This document describes how to start, stop, manage, and troubleshoot HAProxy,
8as well as some known limitations and traps to avoid. It does not describe how
9to configure it (for this please read configuration.txt).
10
11Note to documentation contributors :
12 This document is formatted with 80 columns per line, with even number of
13 spaces for indentation and without tabs. Please follow these rules strictly
14 so that it remains easily printable everywhere. If you add sections, please
15 update the summary below for easier searching.
16
17
18Summary
19-------
20
211. Prerequisites
222. Quick reminder about HAProxy's architecture
233. Starting HAProxy
244. Stopping and restarting HAProxy
255. File-descriptor limitations
266. Memory management
277. CPU usage
288. Logging
299. Statistics and monitoring
Willy Tarreau44aed902015-10-13 14:45:29 +0200309.1. CSV format
Willy Tarreau5d8b9792016-03-11 11:09:34 +0100319.2. Typed output format
329.3. Unix Socket commands
William Lallemand142db372018-12-11 18:56:45 +0100339.4. Master CLI
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003410. Tricks for easier configuration management
3511. Well-known traps to avoid
3612. Debugging and performance issues
3713. Security considerations
38
39
401. Prerequisites
41----------------
42
43In this document it is assumed that the reader has sufficient administration
44skills on a UNIX-like operating system, uses the shell on a daily basis and is
45familiar with troubleshooting utilities such as strace and tcpdump.
46
47
482. Quick reminder about HAProxy's architecture
49----------------------------------------------
50
Willy Tarreau3f364482019-02-27 15:01:46 +010051HAProxy is a multi-threaded, event-driven, non-blocking daemon. This means is
Willy Tarreau2212e6a2015-10-13 14:40:55 +020052uses event multiplexing to schedule all of its activities instead of relying on
53the system to schedule between multiple activities. Most of the time it runs as
54a single process, so the output of "ps aux" on a system will report only one
55"haproxy" process, unless a soft reload is in progress and an older process is
56finishing its job in parallel to the new one. It is thus always easy to trace
Willy Tarreau3f364482019-02-27 15:01:46 +010057its activity using the strace utility. In order to scale with the number of
58available processors, by default haproxy will start one worker thread per
59processor it is allowed to run on. Unless explicitly configured differently,
60the incoming traffic is spread over all these threads, all running the same
61event loop. A great care is taken to limit inter-thread dependencies to the
62strict minimum, so as to try to achieve near-linear scalability. This has some
63impacts such as the fact that a given connection is served by a single thread.
64Thus in order to use all available processing capacity, it is needed to have at
65least as many connections as there are threads, which is almost always granted.
Willy Tarreau2212e6a2015-10-13 14:40:55 +020066
67HAProxy is designed to isolate itself into a chroot jail during startup, where
68it cannot perform any file-system access at all. This is also true for the
69libraries it depends on (eg: libc, libssl, etc). The immediate effect is that
70a running process will not be able to reload a configuration file to apply
71changes, instead a new process will be started using the updated configuration
72file. Some other less obvious effects are that some timezone files or resolver
73files the libc might attempt to access at run time will not be found, though
74this should generally not happen as they're not needed after startup. A nice
75consequence of this principle is that the HAProxy process is totally stateless,
76and no cleanup is needed after it's killed, so any killing method that works
77will do the right thing.
78
79HAProxy doesn't write log files, but it relies on the standard syslog protocol
80to send logs to a remote server (which is often located on the same system).
81
82HAProxy uses its internal clock to enforce timeouts, that is derived from the
83system's time but where unexpected drift is corrected. This is done by limiting
84the time spent waiting in poll() for an event, and measuring the time it really
85took. In practice it never waits more than one second. This explains why, when
86running strace over a completely idle process, periodic calls to poll() (or any
87of its variants) surrounded by two gettimeofday() calls are noticed. They are
88normal, completely harmless and so cheap that the load they imply is totally
89undetectable at the system scale, so there's nothing abnormal there. Example :
90
91 16:35:40.002320 gettimeofday({1442759740, 2605}, NULL) = 0
92 16:35:40.002942 epoll_wait(0, {}, 200, 1000) = 0
93 16:35:41.007542 gettimeofday({1442759741, 7641}, NULL) = 0
94 16:35:41.007998 gettimeofday({1442759741, 8114}, NULL) = 0
95 16:35:41.008391 epoll_wait(0, {}, 200, 1000) = 0
96 16:35:42.011313 gettimeofday({1442759742, 11411}, NULL) = 0
97
98HAProxy is a TCP proxy, not a router. It deals with established connections that
99have been validated by the kernel, and not with packets of any form nor with
100sockets in other states (eg: no SYN_RECV nor TIME_WAIT), though their existence
101may prevent it from binding a port. It relies on the system to accept incoming
102connections and to initiate outgoing connections. An immediate effect of this is
103that there is no relation between packets observed on the two sides of a
104forwarded connection, which can be of different size, numbers and even family.
105Since a connection may only be accepted from a socket in LISTEN state, all the
106sockets it is listening to are necessarily visible using the "netstat" utility
107to show listening sockets. Example :
108
109 # netstat -ltnp
110 Active Internet connections (only servers)
111 Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
112 tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 1629/sshd
113 tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN 2847/haproxy
114 tcp 0 0 0.0.0.0:443 0.0.0.0:* LISTEN 2847/haproxy
115
116
1173. Starting HAProxy
118-------------------
119
120HAProxy is started by invoking the "haproxy" program with a number of arguments
121passed on the command line. The actual syntax is :
122
123 $ haproxy [<options>]*
124
125where [<options>]* is any number of options. An option always starts with '-'
126followed by one of more letters, and possibly followed by one or multiple extra
127arguments. Without any option, HAProxy displays the help page with a reminder
128about supported options. Available options may vary slightly based on the
129operating system. A fair number of these options overlap with an equivalent one
130if the "global" section. In this case, the command line always has precedence
131over the configuration file, so that the command line can be used to quickly
132enforce some settings without touching the configuration files. The current
133list of options is :
134
135 -- <cfgfile>* : all the arguments following "--" are paths to configuration
Maxime de Roucy379d9c72016-05-13 23:52:56 +0200136 file/directory to be loaded and processed in the declaration order. It is
137 mostly useful when relying on the shell to load many files that are
138 numerically ordered. See also "-f". The difference between "--" and "-f" is
139 that one "-f" must be placed before each file name, while a single "--" is
140 needed before all file names. Both options can be used together, the
141 command line ordering still applies. When more than one file is specified,
142 each file must start on a section boundary, so the first keyword of each
143 file must be one of "global", "defaults", "peers", "listen", "frontend",
144 "backend", and so on. A file cannot contain just a server list for example.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200145
Maxime de Roucy379d9c72016-05-13 23:52:56 +0200146 -f <cfgfile|cfgdir> : adds <cfgfile> to the list of configuration files to be
147 loaded. If <cfgdir> is a directory, all the files (and only files) it
Dan Lloyd8e48b872016-07-01 21:01:18 -0400148 contains are added in lexical order (using LC_COLLATE=C) to the list of
Maxime de Roucy379d9c72016-05-13 23:52:56 +0200149 configuration files to be loaded ; only files with ".cfg" extension are
150 added, only non hidden files (not prefixed with ".") are added.
151 Configuration files are loaded and processed in their declaration order.
152 This option may be specified multiple times to load multiple files. See
153 also "--". The difference between "--" and "-f" is that one "-f" must be
154 placed before each file name, while a single "--" is needed before all file
155 names. Both options can be used together, the command line ordering still
156 applies. When more than one file is specified, each file must start on a
157 section boundary, so the first keyword of each file must be one of
158 "global", "defaults", "peers", "listen", "frontend", "backend", and so on.
159 A file cannot contain just a server list for example.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200160
161 -C <dir> : changes to directory <dir> before loading configuration
162 files. This is useful when using relative paths. Warning when using
163 wildcards after "--" which are in fact replaced by the shell before
164 starting haproxy.
165
166 -D : start as a daemon. The process detaches from the current terminal after
167 forking, and errors are not reported anymore in the terminal. It is
168 equivalent to the "daemon" keyword in the "global" section of the
169 configuration. It is recommended to always force it in any init script so
170 that a faulty configuration doesn't prevent the system from booting.
171
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200172 -L <name> : change the local peer name to <name>, which defaults to the local
William Lallemanddaf4cd22018-04-17 16:46:13 +0200173 hostname. This is used only with peers replication. You can use the
174 variable $HAPROXY_LOCALPEER in the configuration file to reference the
175 peer name.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200176
177 -N <limit> : sets the default per-proxy maxconn to <limit> instead of the
178 builtin default value (usually 2000). Only useful for debugging.
179
180 -V : enable verbose mode (disables quiet mode). Reverts the effect of "-q" or
181 "quiet".
182
William Lallemande202b1e2017-06-01 17:38:56 +0200183 -W : master-worker mode. It is equivalent to the "master-worker" keyword in
184 the "global" section of the configuration. This mode will launch a "master"
185 which will monitor the "workers". Using this mode, you can reload HAProxy
186 directly by sending a SIGUSR2 signal to the master. The master-worker mode
187 is compatible either with the foreground or daemon mode. It is
188 recommended to use this mode with multiprocess and systemd.
189
Pavlos Parissisf65f2572018-02-07 21:42:16 +0100190 -Ws : master-worker mode with support of `notify` type of systemd service.
191 This option is only available when HAProxy was built with `USE_SYSTEMD`
192 build option enabled.
193
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200194 -c : only performs a check of the configuration files and exits before trying
195 to bind. The exit status is zero if everything is OK, or non-zero if an
196 error is encountered.
197
198 -d : enable debug mode. This disables daemon mode, forces the process to stay
199 in foreground and to show incoming and outgoing events. It is equivalent to
200 the "global" section's "debug" keyword. It must never be used in an init
201 script.
202
203 -dG : disable use of getaddrinfo() to resolve host names into addresses. It
204 can be used when suspecting that getaddrinfo() doesn't work as expected.
205 This option was made available because many bogus implementations of
206 getaddrinfo() exist on various systems and cause anomalies that are
207 difficult to troubleshoot.
208
Dan Lloyd8e48b872016-07-01 21:01:18 -0400209 -dM[<byte>] : forces memory poisoning, which means that each and every
Willy Tarreaubafbe012017-11-24 17:34:44 +0100210 memory region allocated with malloc() or pool_alloc() will be filled with
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200211 <byte> before being passed to the caller. When <byte> is not specified, it
212 defaults to 0x50 ('P'). While this slightly slows down operations, it is
213 useful to reliably trigger issues resulting from missing initializations in
214 the code that cause random crashes. Note that -dM0 has the effect of
215 turning any malloc() into a calloc(). In any case if a bug appears or
216 disappears when using this option it means there is a bug in haproxy, so
217 please report it.
218
219 -dS : disable use of the splice() system call. It is equivalent to the
220 "global" section's "nosplice" keyword. This may be used when splice() is
221 suspected to behave improperly or to cause performance issues, or when
222 using strace to see the forwarded data (which do not appear when using
223 splice()).
224
225 -dV : disable SSL verify on the server side. It is equivalent to having
226 "ssl-server-verify none" in the "global" section. This is useful when
227 trying to reproduce production issues out of the production
228 environment. Never use this in an init script as it degrades SSL security
229 to the servers.
230
231 -db : disable background mode and multi-process mode. The process remains in
232 foreground. It is mainly used during development or during small tests, as
233 Ctrl-C is enough to stop the process. Never use it in an init script.
234
235 -de : disable the use of the "epoll" poller. It is equivalent to the "global"
236 section's keyword "noepoll". It is mostly useful when suspecting a bug
237 related to this poller. On systems supporting epoll, the fallback will
238 generally be the "poll" poller.
239
240 -dk : disable the use of the "kqueue" poller. It is equivalent to the
241 "global" section's keyword "nokqueue". It is mostly useful when suspecting
242 a bug related to this poller. On systems supporting kqueue, the fallback
243 will generally be the "poll" poller.
244
245 -dp : disable the use of the "poll" poller. It is equivalent to the "global"
246 section's keyword "nopoll". It is mostly useful when suspecting a bug
247 related to this poller. On systems supporting poll, the fallback will
248 generally be the "select" poller, which cannot be disabled and is limited
249 to 1024 file descriptors.
250
Willy Tarreau3eed10e2016-11-07 21:03:16 +0100251 -dr : ignore server address resolution failures. It is very common when
252 validating a configuration out of production not to have access to the same
253 resolvers and to fail on server address resolution, making it difficult to
254 test a configuration. This option simply appends the "none" method to the
255 list of address resolution methods for all servers, ensuring that even if
256 the libc fails to resolve an address, the startup sequence is not
257 interrupted.
258
Willy Tarreau70060452015-12-14 12:46:07 +0100259 -m <limit> : limit the total allocatable memory to <limit> megabytes across
260 all processes. This may cause some connection refusals or some slowdowns
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200261 depending on the amount of memory needed for normal operations. This is
Willy Tarreau70060452015-12-14 12:46:07 +0100262 mostly used to force the processes to work in a constrained resource usage
263 scenario. It is important to note that the memory is not shared between
264 processes, so in a multi-process scenario, this value is first divided by
265 global.nbproc before forking.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200266
267 -n <limit> : limits the per-process connection limit to <limit>. This is
268 equivalent to the global section's keyword "maxconn". It has precedence
269 over this keyword. This may be used to quickly force lower limits to avoid
270 a service outage on systems where resource limits are too low.
271
272 -p <file> : write all processes' pids into <file> during startup. This is
273 equivalent to the "global" section's keyword "pidfile". The file is opened
274 before entering the chroot jail, and after doing the chdir() implied by
275 "-C". Each pid appears on its own line.
276
277 -q : set "quiet" mode. This disables some messages during the configuration
278 parsing and during startup. It can be used in combination with "-c" to
279 just check if a configuration file is valid or not.
280
William Lallemand142db372018-12-11 18:56:45 +0100281 -S <bind>[,bind_options...]: in master-worker mode, bind a master CLI, which
282 allows the access to every processes, running or leaving ones.
283 For security reasons, it is recommended to bind the master CLI to a local
284 UNIX socket. The bind options are the same as the keyword "bind" in
285 the configuration file with words separated by commas instead of spaces.
286
287 Note that this socket can't be used to retrieve the listening sockets from
288 an old process during a seamless reload.
289
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200290 -sf <pid>* : send the "finish" signal (SIGUSR1) to older processes after boot
291 completion to ask them to finish what they are doing and to leave. <pid>
292 is a list of pids to signal (one per argument). The list ends on any
293 option starting with a "-". It is not a problem if the list of pids is
294 empty, so that it can be built on the fly based on the result of a command
295 like "pidof" or "pgrep".
296
297 -st <pid>* : send the "terminate" signal (SIGTERM) to older processes after
298 boot completion to terminate them immediately without finishing what they
299 were doing. <pid> is a list of pids to signal (one per argument). The list
300 is ends on any option starting with a "-". It is not a problem if the list
301 of pids is empty, so that it can be built on the fly based on the result of
302 a command like "pidof" or "pgrep".
303
304 -v : report the version and build date.
305
306 -vv : display the version, build options, libraries versions and usable
307 pollers. This output is systematically requested when filing a bug report.
308
Olivier Houchardd33fc3a2017-04-05 22:50:59 +0200309 -x <unix_socket> : connect to the specified socket and try to retrieve any
310 listening sockets from the old process, and use them instead of trying to
311 bind new ones. This is useful to avoid missing any new connection when
William Lallemandf6975e92017-05-26 17:42:10 +0200312 reloading the configuration on Linux. The capability must be enable on the
313 stats socket using "expose-fd listeners" in your configuration.
Olivier Houchardd33fc3a2017-04-05 22:50:59 +0200314
Dan Lloyd8e48b872016-07-01 21:01:18 -0400315A safe way to start HAProxy from an init file consists in forcing the daemon
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200316mode, storing existing pids to a pid file and using this pid file to notify
317older processes to finish before leaving :
318
319 haproxy -f /etc/haproxy.cfg \
320 -D -p /var/run/haproxy.pid -sf $(cat /var/run/haproxy.pid)
321
322When the configuration is split into a few specific files (eg: tcp vs http),
323it is recommended to use the "-f" option :
324
325 haproxy -f /etc/haproxy/global.cfg -f /etc/haproxy/stats.cfg \
326 -f /etc/haproxy/default-tcp.cfg -f /etc/haproxy/tcp.cfg \
327 -f /etc/haproxy/default-http.cfg -f /etc/haproxy/http.cfg \
328 -D -p /var/run/haproxy.pid -sf $(cat /var/run/haproxy.pid)
329
330When an unknown number of files is expected, such as customer-specific files,
331it is recommended to assign them a name starting with a fixed-size sequence
332number and to use "--" to load them, possibly after loading some defaults :
333
334 haproxy -f /etc/haproxy/global.cfg -f /etc/haproxy/stats.cfg \
335 -f /etc/haproxy/default-tcp.cfg -f /etc/haproxy/tcp.cfg \
336 -f /etc/haproxy/default-http.cfg -f /etc/haproxy/http.cfg \
337 -D -p /var/run/haproxy.pid -sf $(cat /var/run/haproxy.pid) \
338 -f /etc/haproxy/default-customers.cfg -- /etc/haproxy/customers/*
339
340Sometimes a failure to start may happen for whatever reason. Then it is
341important to verify if the version of HAProxy you are invoking is the expected
342version and if it supports the features you are expecting (eg: SSL, PCRE,
343compression, Lua, etc). This can be verified using "haproxy -vv". Some
344important information such as certain build options, the target system and
345the versions of the libraries being used are reported there. It is also what
346you will systematically be asked for when posting a bug report :
347
348 $ haproxy -vv
349 HA-Proxy version 1.6-dev7-a088d3-4 2015/10/08
350 Copyright 2000-2015 Willy Tarreau <willy@haproxy.org>
351
352 Build options :
353 TARGET = linux2628
354 CPU = generic
355 CC = gcc
356 CFLAGS = -pg -O0 -g -fno-strict-aliasing -Wdeclaration-after-statement \
357 -DBUFSIZE=8030 -DMAXREWRITE=1030 -DSO_MARK=36 -DTCP_REPAIR=19
358 OPTIONS = USE_ZLIB=1 USE_DLMALLOC=1 USE_OPENSSL=1 USE_LUA=1 USE_PCRE=1
359
360 Default settings :
361 maxconn = 2000, bufsize = 8030, maxrewrite = 1030, maxpollevents = 200
362
363 Encrypted password support via crypt(3): yes
364 Built with zlib version : 1.2.6
365 Compression algorithms supported : identity("identity"), deflate("deflate"), \
366 raw-deflate("deflate"), gzip("gzip")
367 Built with OpenSSL version : OpenSSL 1.0.1o 12 Jun 2015
368 Running on OpenSSL version : OpenSSL 1.0.1o 12 Jun 2015
369 OpenSSL library supports TLS extensions : yes
370 OpenSSL library supports SNI : yes
371 OpenSSL library supports prefer-server-ciphers : yes
372 Built with PCRE version : 8.12 2011-01-15
373 PCRE library supports JIT : no (USE_PCRE_JIT not set)
374 Built with Lua version : Lua 5.3.1
375 Built with transparent proxy support using: IP_TRANSPARENT IP_FREEBIND
376
377 Available polling systems :
378 epoll : pref=300, test result OK
379 poll : pref=200, test result OK
380 select : pref=150, test result OK
381 Total: 3 (3 usable), will use epoll.
382
383The relevant information that many non-developer users can verify here are :
384 - the version : 1.6-dev7-a088d3-4 above means the code is currently at commit
385 ID "a088d3" which is the 4th one after after official version "1.6-dev7".
386 Version 1.6-dev7 would show as "1.6-dev7-8c1ad7". What matters here is in
387 fact "1.6-dev7". This is the 7th development version of what will become
388 version 1.6 in the future. A development version not suitable for use in
389 production (unless you know exactly what you are doing). A stable version
390 will show as a 3-numbers version, such as "1.5.14-16f863", indicating the
391 14th level of fix on top of version 1.5. This is a production-ready version.
392
393 - the release date : 2015/10/08. It is represented in the universal
394 year/month/day format. Here this means August 8th, 2015. Given that stable
395 releases are issued every few months (1-2 months at the beginning, sometimes
396 6 months once the product becomes very stable), if you're seeing an old date
397 here, it means you're probably affected by a number of bugs or security
398 issues that have since been fixed and that it might be worth checking on the
399 official site.
400
401 - build options : they are relevant to people who build their packages
402 themselves, they can explain why things are not behaving as expected. For
403 example the development version above was built for Linux 2.6.28 or later,
Dan Lloyd8e48b872016-07-01 21:01:18 -0400404 targeting a generic CPU (no CPU-specific optimizations), and lacks any
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200405 code optimization (-O0) so it will perform poorly in terms of performance.
406
407 - libraries versions : zlib version is reported as found in the library
408 itself. In general zlib is considered a very stable product and upgrades
409 are almost never needed. OpenSSL reports two versions, the version used at
410 build time and the one being used, as found on the system. These ones may
411 differ by the last letter but never by the numbers. The build date is also
412 reported because most OpenSSL bugs are security issues and need to be taken
413 seriously, so this library absolutely needs to be kept up to date. Seeing a
414 4-months old version here is highly suspicious and indeed an update was
415 missed. PCRE provides very fast regular expressions and is highly
416 recommended. Certain of its extensions such as JIT are not present in all
417 versions and still young so some people prefer not to build with them,
Dan Lloyd8e48b872016-07-01 21:01:18 -0400418 which is why the build status is reported as well. Regarding the Lua
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200419 scripting language, HAProxy expects version 5.3 which is very young since
420 it was released a little time before HAProxy 1.6. It is important to check
421 on the Lua web site if some fixes are proposed for this branch.
422
423 - Available polling systems will affect the process's scalability when
424 dealing with more than about one thousand of concurrent connections. These
425 ones are only available when the correct system was indicated in the TARGET
426 variable during the build. The "epoll" mechanism is highly recommended on
427 Linux, and the kqueue mechanism is highly recommended on BSD. Lacking them
428 will result in poll() or even select() being used, causing a high CPU usage
429 when dealing with a lot of connections.
430
431
4324. Stopping and restarting HAProxy
433----------------------------------
434
435HAProxy supports a graceful and a hard stop. The hard stop is simple, when the
436SIGTERM signal is sent to the haproxy process, it immediately quits and all
437established connections are closed. The graceful stop is triggered when the
438SIGUSR1 signal is sent to the haproxy process. It consists in only unbinding
439from listening ports, but continue to process existing connections until they
440close. Once the last connection is closed, the process leaves.
441
442The hard stop method is used for the "stop" or "restart" actions of the service
443management script. The graceful stop is used for the "reload" action which
444tries to seamlessly reload a new configuration in a new process.
445
446Both of these signals may be sent by the new haproxy process itself during a
447reload or restart, so that they are sent at the latest possible moment and only
448if absolutely required. This is what is performed by the "-st" (hard) and "-sf"
449(graceful) options respectively.
450
William Lallemande202b1e2017-06-01 17:38:56 +0200451In master-worker mode, it is not needed to start a new haproxy process in
452order to reload the configuration. The master process reacts to the SIGUSR2
453signal by reexecuting itself with the -sf parameter followed by the PIDs of
454the workers. The master will then parse the configuration file and fork new
455workers.
456
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200457To understand better how these signals are used, it is important to understand
458the whole restart mechanism.
459
460First, an existing haproxy process is running. The administrator uses a system
461specific command such as "/etc/init.d/haproxy reload" to indicate he wants to
462take the new configuration file into effect. What happens then is the following.
463First, the service script (/etc/init.d/haproxy or equivalent) will verify that
464the configuration file parses correctly using "haproxy -c". After that it will
465try to start haproxy with this configuration file, using "-st" or "-sf".
466
467Then HAProxy tries to bind to all listening ports. If some fatal errors happen
468(eg: address not present on the system, permission denied), the process quits
469with an error. If a socket binding fails because a port is already in use, then
470the process will first send a SIGTTOU signal to all the pids specified in the
471"-st" or "-sf" pid list. This is what is called the "pause" signal. It instructs
472all existing haproxy processes to temporarily stop listening to their ports so
473that the new process can try to bind again. During this time, the old process
474continues to process existing connections. If the binding still fails (because
475for example a port is shared with another daemon), then the new process sends a
476SIGTTIN signal to the old processes to instruct them to resume operations just
477as if nothing happened. The old processes will then restart listening to the
478ports and continue to accept connections. Not that this mechanism is system
Dan Lloyd8e48b872016-07-01 21:01:18 -0400479dependent and some operating systems may not support it in multi-process mode.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200480
481If the new process manages to bind correctly to all ports, then it sends either
482the SIGTERM (hard stop in case of "-st") or the SIGUSR1 (graceful stop in case
483of "-sf") to all processes to notify them that it is now in charge of operations
484and that the old processes will have to leave, either immediately or once they
485have finished their job.
486
487It is important to note that during this timeframe, there are two small windows
488of a few milliseconds each where it is possible that a few connection failures
489will be noticed during high loads. Typically observed failure rates are around
4901 failure during a reload operation every 10000 new connections per second,
491which means that a heavily loaded site running at 30000 new connections per
492second may see about 3 failed connection upon every reload. The two situations
493where this happens are :
494
495 - if the new process fails to bind due to the presence of the old process,
496 it will first have to go through the SIGTTOU+SIGTTIN sequence, which
497 typically lasts about one millisecond for a few tens of frontends, and
498 during which some ports will not be bound to the old process and not yet
499 bound to the new one. HAProxy works around this on systems that support the
500 SO_REUSEPORT socket options, as it allows the new process to bind without
501 first asking the old one to unbind. Most BSD systems have been supporting
502 this almost forever. Linux has been supporting this in version 2.0 and
503 dropped it around 2.2, but some patches were floating around by then. It
504 was reintroduced in kernel 3.9, so if you are observing a connection
Dan Lloyd8e48b872016-07-01 21:01:18 -0400505 failure rate above the one mentioned above, please ensure that your kernel
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200506 is 3.9 or newer, or that relevant patches were backported to your kernel
507 (less likely).
508
509 - when the old processes close the listening ports, the kernel may not always
510 redistribute any pending connection that was remaining in the socket's
511 backlog. Under high loads, a SYN packet may happen just before the socket
512 is closed, and will lead to an RST packet being sent to the client. In some
513 critical environments where even one drop is not acceptable, these ones are
514 sometimes dealt with using firewall rules to block SYN packets during the
515 reload, forcing the client to retransmit. This is totally system-dependent,
516 as some systems might be able to visit other listening queues and avoid
517 this RST. A second case concerns the ACK from the client on a local socket
518 that was in SYN_RECV state just before the close. This ACK will lead to an
519 RST packet while the haproxy process is still not aware of it. This one is
Dan Lloyd8e48b872016-07-01 21:01:18 -0400520 harder to get rid of, though the firewall filtering rules mentioned above
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200521 will work well if applied one second or so before restarting the process.
522
523For the vast majority of users, such drops will never ever happen since they
524don't have enough load to trigger the race conditions. And for most high traffic
525users, the failure rate is still fairly within the noise margin provided that at
526least SO_REUSEPORT is properly supported on their systems.
527
528
5295. File-descriptor limitations
530------------------------------
531
532In order to ensure that all incoming connections will successfully be served,
533HAProxy computes at load time the total number of file descriptors that will be
534needed during the process's life. A regular Unix process is generally granted
5351024 file descriptors by default, and a privileged process can raise this limit
536itself. This is one reason for starting HAProxy as root and letting it adjust
537the limit. The default limit of 1024 file descriptors roughly allow about 500
538concurrent connections to be processed. The computation is based on the global
539maxconn parameter which limits the total number of connections per process, the
540number of listeners, the number of servers which have a health check enabled,
541the agent checks, the peers, the loggers and possibly a few other technical
542requirements. A simple rough estimate of this number consists in simply
543doubling the maxconn value and adding a few tens to get the approximate number
544of file descriptors needed.
545
546Originally HAProxy did not know how to compute this value, and it was necessary
547to pass the value using the "ulimit-n" setting in the global section. This
548explains why even today a lot of configurations are seen with this setting
549present. Unfortunately it was often miscalculated resulting in connection
550failures when approaching maxconn instead of throttling incoming connection
551while waiting for the needed resources. For this reason it is important to
Dan Lloyd8e48b872016-07-01 21:01:18 -0400552remove any vestigial "ulimit-n" setting that can remain from very old versions.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200553
554Raising the number of file descriptors to accept even moderate loads is
555mandatory but comes with some OS-specific adjustments. First, the select()
556polling system is limited to 1024 file descriptors. In fact on Linux it used
557to be capable of handling more but since certain OS ship with excessively
558restrictive SELinux policies forbidding the use of select() with more than
5591024 file descriptors, HAProxy now refuses to start in this case in order to
560avoid any issue at run time. On all supported operating systems, poll() is
561available and will not suffer from this limitation. It is automatically picked
Dan Lloyd8e48b872016-07-01 21:01:18 -0400562so there is nothing to do to get a working configuration. But poll's becomes
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200563very slow when the number of file descriptors increases. While HAProxy does its
564best to limit this performance impact (eg: via the use of the internal file
565descriptor cache and batched processing), a good rule of thumb is that using
566poll() with more than a thousand concurrent connections will use a lot of CPU.
567
568For Linux systems base on kernels 2.6 and above, the epoll() system call will
569be used. It's a much more scalable mechanism relying on callbacks in the kernel
570that guarantee a constant wake up time regardless of the number of registered
571monitored file descriptors. It is automatically used where detected, provided
572that HAProxy had been built for one of the Linux flavors. Its presence and
573support can be verified using "haproxy -vv".
574
575For BSD systems which support it, kqueue() is available as an alternative. It
576is much faster than poll() and even slightly faster than epoll() thanks to its
577batched handling of changes. At least FreeBSD and OpenBSD support it. Just like
578with Linux's epoll(), its support and availability are reported in the output
579of "haproxy -vv".
580
581Having a good poller is one thing, but it is mandatory that the process can
582reach the limits. When HAProxy starts, it immediately sets the new process's
583file descriptor limits and verifies if it succeeds. In case of failure, it
584reports it before forking so that the administrator can see the problem. As
585long as the process is started by as root, there should be no reason for this
586setting to fail. However, it can fail if the process is started by an
587unprivileged user. If there is a compelling reason for *not* starting haproxy
588as root (eg: started by end users, or by a per-application account), then the
589file descriptor limit can be raised by the system administrator for this
590specific user. The effectiveness of the setting can be verified by issuing
591"ulimit -n" from the user's command line. It should reflect the new limit.
592
593Warning: when an unprivileged user's limits are changed in this user's account,
594it is fairly common that these values are only considered when the user logs in
595and not at all in some scripts run at system boot time nor in crontabs. This is
596totally dependent on the operating system, keep in mind to check "ulimit -n"
597before starting haproxy when running this way. The general advice is never to
598start haproxy as an unprivileged user for production purposes. Another good
599reason is that it prevents haproxy from enabling some security protections.
600
601Once it is certain that the system will allow the haproxy process to use the
602requested number of file descriptors, two new system-specific limits may be
603encountered. The first one is the system-wide file descriptor limit, which is
604the total number of file descriptors opened on the system, covering all
605processes. When this limit is reached, accept() or socket() will typically
606return ENFILE. The second one is the per-process hard limit on the number of
607file descriptors, it prevents setrlimit() from being set higher. Both are very
608dependent on the operating system. On Linux, the system limit is set at boot
609based on the amount of memory. It can be changed with the "fs.file-max" sysctl.
610And the per-process hard limit is set to 1048576 by default, but it can be
611changed using the "fs.nr_open" sysctl.
612
613File descriptor limitations may be observed on a running process when they are
614set too low. The strace utility will report that accept() and socket() return
615"-1 EMFILE" when the process's limits have been reached. In this case, simply
616raising the "ulimit-n" value (or removing it) will solve the problem. If these
617system calls return "-1 ENFILE" then it means that the kernel's limits have
618been reached and that something must be done on a system-wide parameter. These
619trouble must absolutely be addressed, as they result in high CPU usage (when
620accept() fails) and failed connections that are generally visible to the user.
621One solution also consists in lowering the global maxconn value to enforce
622serialization, and possibly to disable HTTP keep-alive to force connections
623to be released and reused faster.
624
625
6266. Memory management
627--------------------
628
629HAProxy uses a simple and fast pool-based memory management. Since it relies on
630a small number of different object types, it's much more efficient to pick new
631objects from a pool which already contains objects of the appropriate size than
632to call malloc() for each different size. The pools are organized as a stack or
633LIFO, so that newly allocated objects are taken from recently released objects
634still hot in the CPU caches. Pools of similar sizes are merged together, in
635order to limit memory fragmentation.
636
637By default, since the focus is set on performance, each released object is put
638back into the pool it came from, and allocated objects are never freed since
639they are expected to be reused very soon.
640
641On the CLI, it is possible to check how memory is being used in pools thanks to
642the "show pools" command :
643
644 > show pools
645 Dumping pools usage. Use SIGQUIT to flush them.
Willy Tarreau0a93b642018-10-16 07:58:39 +0200646 - Pool cache_st (16 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9ccc40=03 [SHARED]
647 - Pool pipe (32 bytes) : 5 allocated (160 bytes), 5 used, 0 failures, 2 users, @0x9ccac0=00 [SHARED]
648 - Pool comp_state (48 bytes) : 3 allocated (144 bytes), 3 used, 0 failures, 5 users, @0x9cccc0=04 [SHARED]
649 - Pool filter (64 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 3 users, @0x9ccbc0=02 [SHARED]
650 - Pool vars (80 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 2 users, @0x9ccb40=01 [SHARED]
651 - Pool uniqueid (128 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 2 users, @0x9cd240=15 [SHARED]
652 - Pool task (144 bytes) : 55 allocated (7920 bytes), 55 used, 0 failures, 1 users, @0x9cd040=11 [SHARED]
653 - Pool session (160 bytes) : 1 allocated (160 bytes), 1 used, 0 failures, 1 users, @0x9cd140=13 [SHARED]
654 - Pool h2s (208 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 2 users, @0x9ccec0=08 [SHARED]
655 - Pool h2c (288 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9cce40=07 [SHARED]
656 - Pool spoe_ctx (304 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 2 users, @0x9ccf40=09 [SHARED]
657 - Pool connection (400 bytes) : 2 allocated (800 bytes), 2 used, 0 failures, 1 users, @0x9cd1c0=14 [SHARED]
658 - Pool hdr_idx (416 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9cd340=17 [SHARED]
659 - Pool dns_resolut (480 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9ccdc0=06 [SHARED]
660 - Pool dns_answer_ (576 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9ccd40=05 [SHARED]
661 - Pool stream (960 bytes) : 1 allocated (960 bytes), 1 used, 0 failures, 1 users, @0x9cd0c0=12 [SHARED]
662 - Pool requri (1024 bytes) : 0 allocated (0 bytes), 0 used, 0 failures, 1 users, @0x9cd2c0=16 [SHARED]
663 - Pool buffer (8030 bytes) : 3 allocated (24090 bytes), 2 used, 0 failures, 1 users, @0x9cd3c0=18 [SHARED]
664 - Pool trash (8062 bytes) : 1 allocated (8062 bytes), 1 used, 0 failures, 1 users, @0x9cd440=19
665 Total: 19 pools, 42296 bytes allocated, 34266 used.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200666
667The pool name is only indicative, it's the name of the first object type using
668this pool. The size in parenthesis is the object size for objects in this pool.
669Object sizes are always rounded up to the closest multiple of 16 bytes. The
670number of objects currently allocated and the equivalent number of bytes is
671reported so that it is easy to know which pool is responsible for the highest
672memory usage. The number of objects currently in use is reported as well in the
673"used" field. The difference between "allocated" and "used" corresponds to the
Willy Tarreau0a93b642018-10-16 07:58:39 +0200674objects that have been freed and are available for immediate use. The address
675at the end of the line is the pool's address, and the following number is the
676pool index when it exists, or is reported as -1 if no index was assigned.
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200677
678It is possible to limit the amount of memory allocated per process using the
679"-m" command line option, followed by a number of megabytes. It covers all of
680the process's addressable space, so that includes memory used by some libraries
681as well as the stack, but it is a reliable limit when building a resource
682constrained system. It works the same way as "ulimit -v" on systems which have
683it, or "ulimit -d" for the other ones.
684
685If a memory allocation fails due to the memory limit being reached or because
686the system doesn't have any enough memory, then haproxy will first start to
687free all available objects from all pools before attempting to allocate memory
688again. This mechanism of releasing unused memory can be triggered by sending
689the signal SIGQUIT to the haproxy process. When doing so, the pools state prior
690to the flush will also be reported to stderr when the process runs in
691foreground.
692
693During a reload operation, the process switched to the graceful stop state also
694automatically performs some flushes after releasing any connection so that all
695possible memory is released to save it for the new process.
696
697
6987. CPU usage
699------------
700
701HAProxy normally spends most of its time in the system and a smaller part in
702userland. A finely tuned 3.5 GHz CPU can sustain a rate about 80000 end-to-end
703connection setups and closes per second at 100% CPU on a single core. When one
704core is saturated, typical figures are :
705 - 95% system, 5% user for long TCP connections or large HTTP objects
706 - 85% system and 15% user for short TCP connections or small HTTP objects in
707 close mode
708 - 70% system and 30% user for small HTTP objects in keep-alive mode
709
710The amount of rules processing and regular expressions will increase the user
711land part. The presence of firewall rules, connection tracking, complex routing
712tables in the system will instead increase the system part.
713
714On most systems, the CPU time observed during network transfers can be cut in 4
715parts :
716 - the interrupt part, which concerns all the processing performed upon I/O
717 receipt, before the target process is even known. Typically Rx packets are
718 accounted for in interrupt. On some systems such as Linux where interrupt
719 processing may be deferred to a dedicated thread, it can appear as softirq,
720 and the thread is called ksoftirqd/0 (for CPU 0). The CPU taking care of
721 this load is generally defined by the hardware settings, though in the case
722 of softirq it is often possible to remap the processing to another CPU.
723 This interrupt part will often be perceived as parasitic since it's not
724 associated with any process, but it actually is some processing being done
725 to prepare the work for the process.
726
727 - the system part, which concerns all the processing done using kernel code
728 called from userland. System calls are accounted as system for example. All
729 synchronously delivered Tx packets will be accounted for as system time. If
730 some packets have to be deferred due to queues filling up, they may then be
731 processed in interrupt context later (eg: upon receipt of an ACK opening a
732 TCP window).
733
734 - the user part, which exclusively runs application code in userland. HAProxy
735 runs exclusively in this part, though it makes heavy use of system calls.
736 Rules processing, regular expressions, compression, encryption all add to
737 the user portion of CPU consumption.
738
739 - the idle part, which is what the CPU does when there is nothing to do. For
740 example HAProxy waits for an incoming connection, or waits for some data to
741 leave, meaning the system is waiting for an ACK from the client to push
742 these data.
743
744In practice regarding HAProxy's activity, it is in general reasonably accurate
745(but totally inexact) to consider that interrupt/softirq are caused by Rx
746processing in kernel drivers, that user-land is caused by layer 7 processing
747in HAProxy, and that system time is caused by network processing on the Tx
748path.
749
750Since HAProxy runs around an event loop, it waits for new events using poll()
751(or any alternative) and processes all these events as fast as possible before
752going back to poll() waiting for new events. It measures the time spent waiting
753in poll() compared to the time spent doing processing events. The ratio of
754polling time vs total time is called the "idle" time, it's the amount of time
755spent waiting for something to happen. This ratio is reported in the stats page
756on the "idle" line, or "Idle_pct" on the CLI. When it's close to 100%, it means
757the load is extremely low. When it's close to 0%, it means that there is
758constantly some activity. While it cannot be very accurate on an overloaded
759system due to other processes possibly preempting the CPU from the haproxy
760process, it still provides a good estimate about how HAProxy considers it is
761working : if the load is low and the idle ratio is low as well, it may indicate
762that HAProxy has a lot of work to do, possibly due to very expensive rules that
763have to be processed. Conversely, if HAProxy indicates the idle is close to
764100% while things are slow, it means that it cannot do anything to speed things
765up because it is already waiting for incoming data to process. In the example
766below, haproxy is completely idle :
767
768 $ echo "show info" | socat - /var/run/haproxy.sock | grep ^Idle
769 Idle_pct: 100
770
771When the idle ratio starts to become very low, it is important to tune the
772system and place processes and interrupts correctly to save the most possible
773CPU resources for all tasks. If a firewall is present, it may be worth trying
774to disable it or to tune it to ensure it is not responsible for a large part
775of the performance limitation. It's worth noting that unloading a stateful
776firewall generally reduces both the amount of interrupt/softirq and of system
777usage since such firewalls act both on the Rx and the Tx paths. On Linux,
778unloading the nf_conntrack and ip_conntrack modules will show whether there is
779anything to gain. If so, then the module runs with default settings and you'll
780have to figure how to tune it for better performance. In general this consists
781in considerably increasing the hash table size. On FreeBSD, "pfctl -d" will
782disable the "pf" firewall and its stateful engine at the same time.
783
784If it is observed that a lot of time is spent in interrupt/softirq, it is
785important to ensure that they don't run on the same CPU. Most systems tend to
786pin the tasks on the CPU where they receive the network traffic because for
787certain workloads it improves things. But with heavily network-bound workloads
788it is the opposite as the haproxy process will have to fight against its kernel
789counterpart. Pinning haproxy to one CPU core and the interrupts to another one,
790all sharing the same L3 cache tends to sensibly increase network performance
791because in practice the amount of work for haproxy and the network stack are
792quite close, so they can almost fill an entire CPU each. On Linux this is done
793using taskset (for haproxy) or using cpu-map (from the haproxy config), and the
794interrupts are assigned under /proc/irq. Many network interfaces support
795multiple queues and multiple interrupts. In general it helps to spread them
796across a small number of CPU cores provided they all share the same L3 cache.
797Please always stop irq_balance which always does the worst possible thing on
798such workloads.
799
800For CPU-bound workloads consisting in a lot of SSL traffic or a lot of
801compression, it may be worth using multiple processes dedicated to certain
802tasks, though there is no universal rule here and experimentation will have to
803be performed.
804
805In order to increase the CPU capacity, it is possible to make HAProxy run as
806several processes, using the "nbproc" directive in the global section. There
807are some limitations though :
808 - health checks are run per process, so the target servers will get as many
809 checks as there are running processes ;
810 - maxconn values and queues are per-process so the correct value must be set
811 to avoid overloading the servers ;
812 - outgoing connections should avoid using port ranges to avoid conflicts
813 - stick-tables are per process and are not shared between processes ;
814 - each peers section may only run on a single process at a time ;
815 - the CLI operations will only act on a single process at a time.
816
817With this in mind, it appears that the easiest setup often consists in having
818one first layer running on multiple processes and in charge for the heavy
819processing, passing the traffic to a second layer running in a single process.
820This mechanism is suited to SSL and compression which are the two CPU-heavy
821features. Instances can easily be chained over UNIX sockets (which are cheaper
fengpeiyuancc123c62016-01-15 16:40:53 +0800822than TCP sockets and which do not waste ports), and the proxy protocol which is
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200823useful to pass client information to the next stage. When doing so, it is
824generally a good idea to bind all the single-process tasks to process number 1
825and extra tasks to next processes, as this will make it easier to generate
826similar configurations for different machines.
827
828On Linux versions 3.9 and above, running HAProxy in multi-process mode is much
829more efficient when each process uses a distinct listening socket on the same
830IP:port ; this will make the kernel evenly distribute the load across all
831processes instead of waking them all up. Please check the "process" option of
832the "bind" keyword lines in the configuration manual for more information.
833
834
8358. Logging
836----------
837
838For logging, HAProxy always relies on a syslog server since it does not perform
839any file-system access. The standard way of using it is to send logs over UDP
840to the log server (by default on port 514). Very commonly this is configured to
841127.0.0.1 where the local syslog daemon is running, but it's also used over the
842network to log to a central server. The central server provides additional
843benefits especially in active-active scenarios where it is desirable to keep
844the logs merged in arrival order. HAProxy may also make use of a UNIX socket to
845send its logs to the local syslog daemon, but it is not recommended at all,
846because if the syslog server is restarted while haproxy runs, the socket will
847be replaced and new logs will be lost. Since HAProxy will be isolated inside a
848chroot jail, it will not have the ability to reconnect to the new socket. It
849has also been observed in field that the log buffers in use on UNIX sockets are
850very small and lead to lost messages even at very light loads. But this can be
851fine for testing however.
852
853It is recommended to add the following directive to the "global" section to
854make HAProxy log to the local daemon using facility "local0" :
855
856 log 127.0.0.1:514 local0
857
858and then to add the following one to each "defaults" section or to each frontend
859and backend section :
860
861 log global
862
863This way, all logs will be centralized through the global definition of where
864the log server is.
865
866Some syslog daemons do not listen to UDP traffic by default, so depending on
867the daemon being used, the syntax to enable this will vary :
868
869 - on sysklogd, you need to pass argument "-r" on the daemon's command line
870 so that it listens to a UDP socket for "remote" logs ; note that there is
871 no way to limit it to address 127.0.0.1 so it will also receive logs from
872 remote systems ;
873
874 - on rsyslogd, the following lines must be added to the configuration file :
875
876 $ModLoad imudp
877 $UDPServerAddress *
878 $UDPServerRun 514
879
880 - on syslog-ng, a new source can be created the following way, it then needs
881 to be added as a valid source in one of the "log" directives :
882
883 source s_udp {
884 udp(ip(127.0.0.1) port(514));
885 };
886
887Please consult your syslog daemon's manual for more information. If no logs are
888seen in the system's log files, please consider the following tests :
889
890 - restart haproxy. Each frontend and backend logs one line indicating it's
891 starting. If these logs are received, it means logs are working.
892
893 - run "strace -tt -s100 -etrace=sendmsg -p <haproxy's pid>" and perform some
894 activity that you expect to be logged. You should see the log messages
895 being sent using sendmsg() there. If they don't appear, restart using
896 strace on top of haproxy. If you still see no logs, it definitely means
897 that something is wrong in your configuration.
898
899 - run tcpdump to watch for port 514, for example on the loopback interface if
900 the traffic is being sent locally : "tcpdump -As0 -ni lo port 514". If the
901 packets are seen there, it's the proof they're sent then the syslogd daemon
902 needs to be troubleshooted.
903
904While traffic logs are sent from the frontends (where the incoming connections
905are accepted), backends also need to be able to send logs in order to report a
906server state change consecutive to a health check. Please consult HAProxy's
907configuration manual for more information regarding all possible log settings.
908
Dan Lloyd8e48b872016-07-01 21:01:18 -0400909It is convenient to chose a facility that is not used by other daemons. HAProxy
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200910examples often suggest "local0" for traffic logs and "local1" for admin logs
911because they're never seen in field. A single facility would be enough as well.
912Having separate logs is convenient for log analysis, but it's also important to
913remember that logs may sometimes convey confidential information, and as such
Dan Lloyd8e48b872016-07-01 21:01:18 -0400914they must not be mixed with other logs that may accidentally be handed out to
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200915unauthorized people.
916
917For in-field troubleshooting without impacting the server's capacity too much,
918it is recommended to make use of the "halog" utility provided with HAProxy.
919This is sort of a grep-like utility designed to process HAProxy log files at
920a very fast data rate. Typical figures range between 1 and 2 GB of logs per
921second. It is capable of extracting only certain logs (eg: search for some
922classes of HTTP status codes, connection termination status, search by response
923time ranges, look for errors only), count lines, limit the output to a number
924of lines, and perform some more advanced statistics such as sorting servers
925by response time or error counts, sorting URLs by time or count, sorting client
926addresses by access count, and so on. It is pretty convenient to quickly spot
927anomalies such as a bot looping on the site, and block them.
928
929
9309. Statistics and monitoring
931----------------------------
932
Willy Tarreau44aed902015-10-13 14:45:29 +0200933It is possible to query HAProxy about its status. The most commonly used
934mechanism is the HTTP statistics page. This page also exposes an alternative
935CSV output format for monitoring tools. The same format is provided on the
936Unix socket.
937
938
9399.1. CSV format
940---------------
941
942The statistics may be consulted either from the unix socket or from the HTTP
943page. Both means provide a CSV format whose fields follow. The first line
944begins with a sharp ('#') and has one word per comma-delimited field which
945represents the title of the column. All other lines starting at the second one
946use a classical CSV format using a comma as the delimiter, and the double quote
947('"') as an optional text delimiter, but only if the enclosed text is ambiguous
948(if it contains a quote or a comma). The double-quote character ('"') in the
949text is doubled ('""'), which is the format that most tools recognize. Please
950do not insert any column before these ones in order not to break tools which
951use hard-coded column positions.
952
953In brackets after each field name are the types which may have a value for
954that field. The types are L (Listeners), F (Frontends), B (Backends), and
955S (Servers).
956
957 0. pxname [LFBS]: proxy name
958 1. svname [LFBS]: service name (FRONTEND for frontend, BACKEND for backend,
959 any name for server/listener)
960 2. qcur [..BS]: current queued requests. For the backend this reports the
961 number queued without a server assigned.
962 3. qmax [..BS]: max value of qcur
963 4. scur [LFBS]: current sessions
964 5. smax [LFBS]: max sessions
965 6. slim [LFBS]: configured session limit
Willy Tarreauc73810f2016-01-11 13:52:04 +0100966 7. stot [LFBS]: cumulative number of sessions
Willy Tarreau44aed902015-10-13 14:45:29 +0200967 8. bin [LFBS]: bytes in
968 9. bout [LFBS]: bytes out
969 10. dreq [LFB.]: requests denied because of security concerns.
970 - For tcp this is because of a matched tcp-request content rule.
971 - For http this is because of a matched http-request or tarpit rule.
972 11. dresp [LFBS]: responses denied because of security concerns.
973 - For http this is because of a matched http-request rule, or
974 "option checkcache".
975 12. ereq [LF..]: request errors. Some of the possible causes are:
976 - early termination from the client, before the request has been sent.
977 - read error from the client
978 - client timeout
979 - client closed connection
980 - various bad requests from the client.
981 - request was tarpitted.
982 13. econ [..BS]: number of requests that encountered an error trying to
983 connect to a backend server. The backend stat is the sum of the stat
984 for all servers of that backend, plus any connection errors not
985 associated with a particular server (such as the backend having no
986 active servers).
987 14. eresp [..BS]: response errors. srv_abrt will be counted here also.
988 Some other errors are:
989 - write error on the client socket (won't be counted for the server stat)
990 - failure applying filters to the response.
991 15. wretr [..BS]: number of times a connection to a server was retried.
992 16. wredis [..BS]: number of times a request was redispatched to another
993 server. The server value counts the number of times that server was
994 switched away from.
Willy Tarreaub96dd282016-11-09 14:45:51 +0100995 17. status [LFBS]: status (UP/DOWN/NOLB/MAINT/MAINT(via)/MAINT(resolution)...)
Willy Tarreau44aed902015-10-13 14:45:29 +0200996 18. weight [..BS]: total weight (backend), server weight (server)
997 19. act [..BS]: number of active servers (backend), server is active (server)
998 20. bck [..BS]: number of backup servers (backend), server is backup (server)
999 21. chkfail [...S]: number of failed checks. (Only counts checks failed when
1000 the server is up.)
1001 22. chkdown [..BS]: number of UP->DOWN transitions. The backend counter counts
1002 transitions to the whole backend being down, rather than the sum of the
1003 counters for each server.
1004 23. lastchg [..BS]: number of seconds since the last UP<->DOWN transition
1005 24. downtime [..BS]: total downtime (in seconds). The value for the backend
1006 is the downtime for the whole backend, not the sum of the server downtime.
1007 25. qlimit [...S]: configured maxqueue for the server, or nothing in the
1008 value is 0 (default, meaning no limit)
1009 26. pid [LFBS]: process id (0 for first instance, 1 for second, ...)
1010 27. iid [LFBS]: unique proxy id
1011 28. sid [L..S]: server id (unique inside a proxy)
1012 29. throttle [...S]: current throttle percentage for the server, when
1013 slowstart is active, or no value if not in slowstart.
1014 30. lbtot [..BS]: total number of times a server was selected, either for new
1015 sessions, or when re-dispatching. The server counter is the number
1016 of times that server was selected.
1017 31. tracked [...S]: id of proxy/server if tracking is enabled.
1018 32. type [LFBS]: (0=frontend, 1=backend, 2=server, 3=socket/listener)
1019 33. rate [.FBS]: number of sessions per second over last elapsed second
1020 34. rate_lim [.F..]: configured limit on new sessions per second
1021 35. rate_max [.FBS]: max number of new sessions per second
1022 36. check_status [...S]: status of last health check, one of:
1023 UNK -> unknown
1024 INI -> initializing
1025 SOCKERR -> socket error
1026 L4OK -> check passed on layer 4, no upper layers testing enabled
1027 L4TOUT -> layer 1-4 timeout
1028 L4CON -> layer 1-4 connection problem, for example
1029 "Connection refused" (tcp rst) or "No route to host" (icmp)
1030 L6OK -> check passed on layer 6
1031 L6TOUT -> layer 6 (SSL) timeout
1032 L6RSP -> layer 6 invalid response - protocol error
1033 L7OK -> check passed on layer 7
1034 L7OKC -> check conditionally passed on layer 7, for example 404 with
1035 disable-on-404
1036 L7TOUT -> layer 7 (HTTP/SMTP) timeout
1037 L7RSP -> layer 7 invalid response - protocol error
1038 L7STS -> layer 7 response error, for example HTTP 5xx
Daniel Schnellerb6c8b0d2017-09-01 19:13:55 +02001039 Notice: If a check is currently running, the last known status will be
1040 reported, prefixed with "* ". e. g. "* L7OK".
Willy Tarreau44aed902015-10-13 14:45:29 +02001041 37. check_code [...S]: layer5-7 code, if available
1042 38. check_duration [...S]: time in ms took to finish last health check
1043 39. hrsp_1xx [.FBS]: http responses with 1xx code
1044 40. hrsp_2xx [.FBS]: http responses with 2xx code
1045 41. hrsp_3xx [.FBS]: http responses with 3xx code
1046 42. hrsp_4xx [.FBS]: http responses with 4xx code
1047 43. hrsp_5xx [.FBS]: http responses with 5xx code
1048 44. hrsp_other [.FBS]: http responses with other codes (protocol error)
1049 45. hanafail [...S]: failed health checks details
1050 46. req_rate [.F..]: HTTP requests per second over last elapsed second
1051 47. req_rate_max [.F..]: max number of HTTP requests per second observed
Willy Tarreaufb981bd2016-12-12 14:31:46 +01001052 48. req_tot [.FB.]: total number of HTTP requests received
Willy Tarreau44aed902015-10-13 14:45:29 +02001053 49. cli_abrt [..BS]: number of data transfers aborted by the client
1054 50. srv_abrt [..BS]: number of data transfers aborted by the server
1055 (inc. in eresp)
1056 51. comp_in [.FB.]: number of HTTP response bytes fed to the compressor
1057 52. comp_out [.FB.]: number of HTTP response bytes emitted by the compressor
1058 53. comp_byp [.FB.]: number of bytes that bypassed the HTTP compressor
1059 (CPU/BW limit)
1060 54. comp_rsp [.FB.]: number of HTTP responses that were compressed
1061 55. lastsess [..BS]: number of seconds since last session assigned to
1062 server/backend
1063 56. last_chk [...S]: last health check contents or textual error
1064 57. last_agt [...S]: last agent check contents or textual error
1065 58. qtime [..BS]: the average queue time in ms over the 1024 last requests
1066 59. ctime [..BS]: the average connect time in ms over the 1024 last requests
1067 60. rtime [..BS]: the average response time in ms over the 1024 last requests
1068 (0 for TCP)
1069 61. ttime [..BS]: the average total session time in ms over the 1024 last
1070 requests
Willy Tarreau7f618842016-01-08 11:40:03 +01001071 62. agent_status [...S]: status of last agent check, one of:
1072 UNK -> unknown
1073 INI -> initializing
1074 SOCKERR -> socket error
1075 L4OK -> check passed on layer 4, no upper layers testing enabled
1076 L4TOUT -> layer 1-4 timeout
1077 L4CON -> layer 1-4 connection problem, for example
1078 "Connection refused" (tcp rst) or "No route to host" (icmp)
1079 L7OK -> agent reported "up"
1080 L7STS -> agent reported "fail", "stop", or "down"
1081 63. agent_code [...S]: numeric code reported by agent if any (unused for now)
1082 64. agent_duration [...S]: time in ms taken to finish last check
Willy Tarreaudd7354b2016-01-08 13:47:26 +01001083 65. check_desc [...S]: short human-readable description of check_status
1084 66. agent_desc [...S]: short human-readable description of agent_status
Willy Tarreau3141f592016-01-08 14:25:28 +01001085 67. check_rise [...S]: server's "rise" parameter used by checks
1086 68. check_fall [...S]: server's "fall" parameter used by checks
1087 69. check_health [...S]: server's health check value between 0 and rise+fall-1
1088 70. agent_rise [...S]: agent's "rise" parameter, normally 1
1089 71. agent_fall [...S]: agent's "fall" parameter, normally 1
1090 72. agent_health [...S]: agent's health parameter, between 0 and rise+fall-1
Willy Tarreaua6f5a732016-01-08 16:59:56 +01001091 73. addr [L..S]: address:port or "unix". IPv6 has brackets around the address.
Willy Tarreaue4847c62016-01-08 15:43:54 +01001092 74: cookie [..BS]: server's cookie value or backend's cookie name
Willy Tarreauf8211df2016-01-11 14:09:38 +01001093 75: mode [LFBS]: proxy mode (tcp, http, health, unknown)
Willy Tarreauf1516d92016-01-11 14:48:36 +01001094 76: algo [..B.]: load balancing algorithm
Willy Tarreauc73810f2016-01-11 13:52:04 +01001095 77: conn_rate [.F..]: number of connections over the last elapsed second
1096 78: conn_rate_max [.F..]: highest known conn_rate
1097 79: conn_tot [.F..]: cumulative number of connections
Willy Tarreau5b9bdff2016-01-11 14:40:47 +01001098 80: intercepted [.FB.]: cum. number of intercepted requests (monitor, stats)
Willy Tarreau8a90b8e2016-10-21 18:15:32 +02001099 81: dcon [LF..]: requests denied by "tcp-request connection" rules
Willy Tarreaua5bc36b2016-10-21 18:16:27 +02001100 82: dses [LF..]: requests denied by "tcp-request session" rules
Willy Tarreauea96a822018-05-28 15:15:43 +02001101 83: wrew [LFBS]: cumulative number of failed header rewriting warnings
Jérôme Magnin708eb882019-07-17 09:24:46 +02001102 84: connect [..BS]: cumulative number of connection establishment attempts
1103 85: reuse [..BS]: cumulative number of connection reuses
Willy Tarreau72974292019-11-08 07:29:34 +01001104 86: cache_lookups [.FB.]: cumulative number of cache lookups
Jérôme Magnin34ebb5c2019-07-17 14:04:40 +02001105 87: cache_hits [.FB.]: cumulative number of cache hits
Christopher Faulet2ac25742019-11-08 15:27:27 +01001106 88: srv_icur [...S]: current number of idle connections available for reuse
1107 89: src_ilim [...S]: limit on the number of available idle connections
1108 90. qtime_max [..BS]: the maximum observed queue time in ms
1109 91. ctime_max [..BS]: the maximum observed connect time in ms
1110 92. rtime_max [..BS]: the maximum observed response time in ms (0 for TCP)
1111 93. ttime_max [..BS]: the maximum observed total session time in ms
Christopher Faulet0159ee42019-12-16 14:40:39 +01001112 94. eint [LFBS]: cumulative number of internal errors
Willy Tarreau44aed902015-10-13 14:45:29 +02001113
1114
Willy Tarreau5d8b9792016-03-11 11:09:34 +010011159.2) Typed output format
1116------------------------
1117
1118Both "show info" and "show stat" support a mode where each output value comes
1119with its type and sufficient information to know how the value is supposed to
1120be aggregated between processes and how it evolves.
1121
1122In all cases, the output consists in having a single value per line with all
1123the information split into fields delimited by colons (':').
1124
1125The first column designates the object or metric being dumped. Its format is
1126specific to the command producing this output and will not be described in this
1127section. Usually it will consist in a series of identifiers and field names.
1128
1129The second column contains 3 characters respectively indicating the origin, the
1130nature and the scope of the value being reported. The first character (the
1131origin) indicates where the value was extracted from. Possible characters are :
1132
1133 M The value is a metric. It is valid at one instant any may change depending
1134 on its nature .
1135
1136 S The value is a status. It represents a discrete value which by definition
1137 cannot be aggregated. It may be the status of a server ("UP" or "DOWN"),
1138 the PID of the process, etc.
1139
1140 K The value is a sorting key. It represents an identifier which may be used
1141 to group some values together because it is unique among its class. All
1142 internal identifiers are keys. Some names can be listed as keys if they
1143 are unique (eg: a frontend name is unique). In general keys come from the
Dan Lloyd8e48b872016-07-01 21:01:18 -04001144 configuration, even though some of them may automatically be assigned. For
Willy Tarreau5d8b9792016-03-11 11:09:34 +01001145 most purposes keys may be considered as equivalent to configuration.
1146
1147 C The value comes from the configuration. Certain configuration values make
1148 sense on the output, for example a concurrent connection limit or a cookie
1149 name. By definition these values are the same in all processes started
1150 from the same configuration file.
1151
1152 P The value comes from the product itself. There are very few such values,
1153 most common use is to report the product name, version and release date.
1154 These elements are also the same between all processes.
1155
1156The second character (the nature) indicates the nature of the information
1157carried by the field in order to let an aggregator decide on what operation to
1158use to aggregate multiple values. Possible characters are :
1159
1160 A The value represents an age since a last event. This is a bit different
1161 from the duration in that an age is automatically computed based on the
1162 current date. A typical example is how long ago did the last session
1163 happen on a server. Ages are generally aggregated by taking the minimum
1164 value and do not need to be stored.
1165
1166 a The value represents an already averaged value. The average response times
1167 and server weights are of this nature. Averages can typically be averaged
1168 between processes.
1169
1170 C The value represents a cumulative counter. Such measures perpetually
1171 increase until they wrap around. Some monitoring protocols need to tell
1172 the difference between a counter and a gauge to report a different type.
1173 In general counters may simply be summed since they represent events or
1174 volumes. Examples of metrics of this nature are connection counts or byte
1175 counts.
1176
1177 D The value represents a duration for a status. There are a few usages of
1178 this, most of them include the time taken by the last health check and
1179 the time a server has spent down. Durations are generally not summed,
1180 most of the time the maximum will be retained to compute an SLA.
1181
1182 G The value represents a gauge. It's a measure at one instant. The memory
1183 usage or the current number of active connections are of this nature.
1184 Metrics of this type are typically summed during aggregation.
1185
1186 L The value represents a limit (generally a configured one). By nature,
1187 limits are harder to aggregate since they are specific to the point where
1188 they were retrieved. In certain situations they may be summed or be kept
1189 separate.
1190
1191 M The value represents a maximum. In general it will apply to a gauge and
1192 keep the highest known value. An example of such a metric could be the
1193 maximum amount of concurrent connections that was encountered in the
1194 product's life time. To correctly aggregate maxima, you are supposed to
1195 output a range going from the maximum of all maxima and the sum of all
1196 of them. There is indeed no way to know if they were encountered
1197 simultaneously or not.
1198
1199 m The value represents a minimum. In general it will apply to a gauge and
1200 keep the lowest known value. An example of such a metric could be the
1201 minimum amount of free memory pools that was encountered in the product's
1202 life time. To correctly aggregate minima, you are supposed to output a
1203 range going from the minimum of all minima and the sum of all of them.
1204 There is indeed no way to know if they were encountered simultaneously
1205 or not.
1206
1207 N The value represents a name, so it is a string. It is used to report
1208 proxy names, server names and cookie names. Names have configuration or
1209 keys as their origin and are supposed to be the same among all processes.
1210
1211 O The value represents a free text output. Outputs from various commands,
1212 returns from health checks, node descriptions are of such nature.
1213
1214 R The value represents an event rate. It's a measure at one instant. It is
1215 quite similar to a gauge except that the recipient knows that this measure
1216 moves slowly and may decide not to keep all values. An example of such a
1217 metric is the measured amount of connections per second. Metrics of this
1218 type are typically summed during aggregation.
1219
1220 T The value represents a date or time. A field emitting the current date
1221 would be of this type. The method to aggregate such information is left
1222 as an implementation choice. For now no field uses this type.
1223
1224The third character (the scope) indicates what extent the value reflects. Some
1225elements may be per process while others may be per configuration or per system.
1226The distinction is important to know whether or not a single value should be
1227kept during aggregation or if values have to be aggregated. The following
1228characters are currently supported :
1229
1230 C The value is valid for a whole cluster of nodes, which is the set of nodes
1231 communicating over the peers protocol. An example could be the amount of
1232 entries present in a stick table that is replicated with other peers. At
1233 the moment no metric use this scope.
1234
1235 P The value is valid only for the process reporting it. Most metrics use
1236 this scope.
1237
1238 S The value is valid for the whole service, which is the set of processes
1239 started together from the same configuration file. All metrics originating
1240 from the configuration use this scope. Some other metrics may use it as
1241 well for some shared resources (eg: shared SSL cache statistics).
1242
1243 s The value is valid for the whole system, such as the system's hostname,
1244 current date or resource usage. At the moment this scope is not used by
1245 any metric.
1246
1247Consumers of these information will generally have enough of these 3 characters
1248to determine how to accurately report aggregated information across multiple
1249processes.
1250
1251After this column, the third column indicates the type of the field, among "s32"
1252(signed 32-bit integer), "s64" (signed 64-bit integer), "u32" (unsigned 32-bit
1253integer), "u64" (unsigned 64-bit integer), "str" (string). It is important to
1254know the type before parsing the value in order to properly read it. For example
1255a string containing only digits is still a string an not an integer (eg: an
1256error code extracted by a check).
1257
1258Then the fourth column is the value itself, encoded according to its type.
1259Strings are dumped as-is immediately after the colon without any leading space.
1260If a string contains a colon, it will appear normally. This means that the
1261output should not be exclusively split around colons or some check outputs
1262or server addresses might be truncated.
1263
1264
12659.3. Unix Socket commands
Willy Tarreau44aed902015-10-13 14:45:29 +02001266-------------------------
1267
1268The stats socket is not enabled by default. In order to enable it, it is
1269necessary to add one line in the global section of the haproxy configuration.
1270A second line is recommended to set a larger timeout, always appreciated when
1271issuing commands by hand :
1272
1273 global
1274 stats socket /var/run/haproxy.sock mode 600 level admin
1275 stats timeout 2m
1276
1277It is also possible to add multiple instances of the stats socket by repeating
1278the line, and make them listen to a TCP port instead of a UNIX socket. This is
1279never done by default because this is dangerous, but can be handy in some
1280situations :
1281
1282 global
1283 stats socket /var/run/haproxy.sock mode 600 level admin
1284 stats socket ipv4@192.168.0.1:9999 level admin
1285 stats timeout 2m
1286
1287To access the socket, an external utility such as "socat" is required. Socat is
1288a swiss-army knife to connect anything to anything. We use it to connect
1289terminals to the socket, or a couple of stdin/stdout pipes to it for scripts.
1290The two main syntaxes we'll use are the following :
1291
1292 # socat /var/run/haproxy.sock stdio
1293 # socat /var/run/haproxy.sock readline
1294
1295The first one is used with scripts. It is possible to send the output of a
1296script to haproxy, and pass haproxy's output to another script. That's useful
1297for retrieving counters or attack traces for example.
1298
1299The second one is only useful for issuing commands by hand. It has the benefit
1300that the terminal is handled by the readline library which supports line
1301editing and history, which is very convenient when issuing repeated commands
1302(eg: watch a counter).
1303
1304The socket supports two operation modes :
1305 - interactive
1306 - non-interactive
1307
1308The non-interactive mode is the default when socat connects to the socket. In
1309this mode, a single line may be sent. It is processed as a whole, responses are
1310sent back, and the connection closes after the end of the response. This is the
1311mode that scripts and monitoring tools use. It is possible to send multiple
1312commands in this mode, they need to be delimited by a semi-colon (';'). For
1313example :
1314
1315 # echo "show info;show stat;show table" | socat /var/run/haproxy stdio
1316
Dragan Dosena1c35ab2016-11-24 11:33:12 +01001317If a command needs to use a semi-colon or a backslash (eg: in a value), it
Joseph Herlant71b4b152018-11-13 16:55:16 -08001318must be preceded by a backslash ('\').
Chad Lavoiee3f50312016-05-26 16:42:25 -04001319
Willy Tarreau44aed902015-10-13 14:45:29 +02001320The interactive mode displays a prompt ('>') and waits for commands to be
1321entered on the line, then processes them, and displays the prompt again to wait
1322for a new command. This mode is entered via the "prompt" command which must be
1323sent on the first line in non-interactive mode. The mode is a flip switch, if
1324"prompt" is sent in interactive mode, it is disabled and the connection closes
1325after processing the last command of the same line.
1326
1327For this reason, when debugging by hand, it's quite common to start with the
1328"prompt" command :
1329
1330 # socat /var/run/haproxy readline
1331 prompt
1332 > show info
1333 ...
1334 >
1335
1336Since multiple commands may be issued at once, haproxy uses the empty line as a
1337delimiter to mark an end of output for each command, and takes care of ensuring
1338that no command can emit an empty line on output. A script can thus easily
1339parse the output even when multiple commands were pipelined on a single line.
1340
Aurélien Nephtaliabbf6072018-04-18 13:26:46 +02001341Some commands may take an optional payload. To add one to a command, the first
1342line needs to end with the "<<\n" pattern. The next lines will be treated as
1343the payload and can contain as many lines as needed. To validate a command with
1344a payload, it needs to end with an empty line.
1345
1346Limitations do exist: the length of the whole buffer passed to the CLI must
1347not be greater than tune.bfsize and the pattern "<<" must not be glued to the
1348last word of the line.
1349
1350When entering a paylod while in interactive mode, the prompt will change from
1351"> " to "+ ".
1352
Willy Tarreau44aed902015-10-13 14:45:29 +02001353It is important to understand that when multiple haproxy processes are started
1354on the same sockets, any process may pick up the request and will output its
1355own stats.
1356
1357The list of commands currently supported on the stats socket is provided below.
1358If an unknown command is sent, haproxy displays the usage message which reminds
1359all supported commands. Some commands support a more complex syntax, generally
1360it will explain what part of the command is invalid when this happens.
1361
Olivier Doucetd8703e82017-08-31 11:05:10 +02001362Some commands require a higher level of privilege to work. If you do not have
1363enough privilege, you will get an error "Permission denied". Please check
1364the "level" option of the "bind" keyword lines in the configuration manual
1365for more information.
1366
William Lallemand6ab08b32019-11-29 16:48:43 +01001367abort ssl cert <filename>
1368 Abort and destroy a temporary SSL certificate update transaction.
1369
1370 See also "set ssl cert" and "commit ssl cert".
1371
Willy Tarreau44aed902015-10-13 14:45:29 +02001372add acl <acl> <pattern>
1373 Add an entry into the acl <acl>. <acl> is the #<id> or the <file> returned by
1374 "show acl". This command does not verify if the entry already exists. This
1375 command cannot be used if the reference <acl> is a file also used with a map.
1376 In this case, you must use the command "add map" in place of "add acl".
1377
1378add map <map> <key> <value>
Aurélien Nephtali25650ce2018-04-18 14:04:47 +02001379add map <map> <payload>
Willy Tarreau44aed902015-10-13 14:45:29 +02001380 Add an entry into the map <map> to associate the value <value> to the key
1381 <key>. This command does not verify if the entry already exists. It is
1382 mainly used to fill a map after a clear operation. Note that if the reference
1383 <map> is a file and is shared with a map, this map will contain also a new
Aurélien Nephtali25650ce2018-04-18 14:04:47 +02001384 pattern entry. Using the payload syntax it is possible to add multiple
1385 key/value pairs by entering them on separate lines. On each new line, the
1386 first word is the key and the rest of the line is considered to be the value
1387 which can even contains spaces.
1388
1389 Example:
1390
1391 # socat /tmp/sock1 -
1392 prompt
1393
1394 > add map #-1 <<
1395 + key1 value1
1396 + key2 value2 with spaces
1397 + key3 value3 also with spaces
1398 + key4 value4
1399
1400 >
Willy Tarreau44aed902015-10-13 14:45:29 +02001401
1402clear counters
1403 Clear the max values of the statistics counters in each proxy (frontend &
Willy Tarreaud80cb4e2018-01-20 19:30:13 +01001404 backend) and in each server. The accumulated counters are not affected. The
1405 internal activity counters reported by "show activity" are also reset. This
Willy Tarreau44aed902015-10-13 14:45:29 +02001406 can be used to get clean counters after an incident, without having to
1407 restart nor to clear traffic counters. This command is restricted and can
1408 only be issued on sockets configured for levels "operator" or "admin".
1409
1410clear counters all
1411 Clear all statistics counters in each proxy (frontend & backend) and in each
1412 server. This has the same effect as restarting. This command is restricted
1413 and can only be issued on sockets configured for level "admin".
1414
1415clear acl <acl>
1416 Remove all entries from the acl <acl>. <acl> is the #<id> or the <file>
1417 returned by "show acl". Note that if the reference <acl> is a file and is
1418 shared with a map, this map will be also cleared.
1419
1420clear map <map>
1421 Remove all entries from the map <map>. <map> is the #<id> or the <file>
1422 returned by "show map". Note that if the reference <map> is a file and is
1423 shared with a acl, this acl will be also cleared.
1424
1425clear table <table> [ data.<type> <operator> <value> ] | [ key <key> ]
1426 Remove entries from the stick-table <table>.
1427
1428 This is typically used to unblock some users complaining they have been
1429 abusively denied access to a service, but this can also be used to clear some
1430 stickiness entries matching a server that is going to be replaced (see "show
1431 table" below for details). Note that sometimes, removal of an entry will be
1432 refused because it is currently tracked by a session. Retrying a few seconds
1433 later after the session ends is usual enough.
1434
1435 In the case where no options arguments are given all entries will be removed.
1436
1437 When the "data." form is used entries matching a filter applied using the
1438 stored data (see "stick-table" in section 4.2) are removed. A stored data
1439 type must be specified in <type>, and this data type must be stored in the
1440 table otherwise an error is reported. The data is compared according to
1441 <operator> with the 64-bit integer <value>. Operators are the same as with
1442 the ACLs :
1443
1444 - eq : match entries whose data is equal to this value
1445 - ne : match entries whose data is not equal to this value
1446 - le : match entries whose data is less than or equal to this value
1447 - ge : match entries whose data is greater than or equal to this value
1448 - lt : match entries whose data is less than this value
1449 - gt : match entries whose data is greater than this value
1450
1451 When the key form is used the entry <key> is removed. The key must be of the
1452 same type as the table, which currently is limited to IPv4, IPv6, integer and
1453 string.
1454
1455 Example :
1456 $ echo "show table http_proxy" | socat stdio /tmp/sock1
1457 >>> # table: http_proxy, type: ip, size:204800, used:2
1458 >>> 0x80e6a4c: key=127.0.0.1 use=0 exp=3594729 gpc0=0 conn_rate(30000)=1 \
1459 bytes_out_rate(60000)=187
1460 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
1461 bytes_out_rate(60000)=191
1462
1463 $ echo "clear table http_proxy key 127.0.0.1" | socat stdio /tmp/sock1
1464
1465 $ echo "show table http_proxy" | socat stdio /tmp/sock1
1466 >>> # table: http_proxy, type: ip, size:204800, used:1
1467 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
1468 bytes_out_rate(60000)=191
1469 $ echo "clear table http_proxy data.gpc0 eq 1" | socat stdio /tmp/sock1
1470 $ echo "show table http_proxy" | socat stdio /tmp/sock1
1471 >>> # table: http_proxy, type: ip, size:204800, used:1
1472
William Lallemand6ab08b32019-11-29 16:48:43 +01001473commit ssl cert <filename>
1474 Commit and apply a temporary SSL certificate update transaction.
1475 Generate every SSL contextes and SNIs it needs, insert them, and remove
1476 the previous ones. Replace in memory the previous SSL certificates
1477 everywhere the <filename> was used in the configuration.
1478 Upon failure it doesn't remove or insert anything. Once the temporary
1479 transaction is commited, it is destroyed.
1480
1481 See also "ssl set cert" and "abort ssl cert".
1482
Willy Tarreau6bdf3e92019-05-20 14:25:05 +02001483debug dev <command> [args]*
Willy Tarreaub24ab222019-10-24 18:03:39 +02001484 Call a developer-specific command. Only supported on a CLI connection running
1485 in expert mode (see "expert-mode on"). Such commands are extremely dangerous
1486 and not forgiving, any misuse may result in a crash of the process. They are
1487 intended for experts only, and must really not be used unless told to do so.
1488 Some of them are only available when haproxy is built with DEBUG_DEV defined
1489 because they may have security implications. All of these commands require
1490 admin privileges, and are purposely not documented to avoid encouraging their
1491 use by people who are not at ease with the source code.
Willy Tarreau6bdf3e92019-05-20 14:25:05 +02001492
Willy Tarreau44aed902015-10-13 14:45:29 +02001493del acl <acl> [<key>|#<ref>]
1494 Delete all the acl entries from the acl <acl> corresponding to the key <key>.
1495 <acl> is the #<id> or the <file> returned by "show acl". If the <ref> is used,
1496 this command delete only the listed reference. The reference can be found with
1497 listing the content of the acl. Note that if the reference <acl> is a file and
1498 is shared with a map, the entry will be also deleted in the map.
1499
1500del map <map> [<key>|#<ref>]
1501 Delete all the map entries from the map <map> corresponding to the key <key>.
1502 <map> is the #<id> or the <file> returned by "show map". If the <ref> is used,
1503 this command delete only the listed reference. The reference can be found with
1504 listing the content of the map. Note that if the reference <map> is a file and
1505 is shared with a acl, the entry will be also deleted in the map.
1506
1507disable agent <backend>/<server>
1508 Mark the auxiliary agent check as temporarily stopped.
1509
1510 In the case where an agent check is being run as a auxiliary check, due
1511 to the agent-check parameter of a server directive, new checks are only
Dan Lloyd8e48b872016-07-01 21:01:18 -04001512 initialized when the agent is in the enabled. Thus, disable agent will
Willy Tarreau44aed902015-10-13 14:45:29 +02001513 prevent any new agent checks from begin initiated until the agent
1514 re-enabled using enable agent.
1515
1516 When an agent is disabled the processing of an auxiliary agent check that
1517 was initiated while the agent was set as enabled is as follows: All
1518 results that would alter the weight, specifically "drain" or a weight
1519 returned by the agent, are ignored. The processing of agent check is
1520 otherwise unchanged.
1521
1522 The motivation for this feature is to allow the weight changing effects
1523 of the agent checks to be paused to allow the weight of a server to be
1524 configured using set weight without being overridden by the agent.
1525
1526 This command is restricted and can only be issued on sockets configured for
1527 level "admin".
1528
Olivier Houchard614f8d72017-03-14 20:08:46 +01001529disable dynamic-cookie backend <backend>
1530 Disable the generation of dynamic cookies fot the backend <backend>
1531
Willy Tarreau44aed902015-10-13 14:45:29 +02001532disable frontend <frontend>
1533 Mark the frontend as temporarily stopped. This corresponds to the mode which
1534 is used during a soft restart : the frontend releases the port but can be
1535 enabled again if needed. This should be used with care as some non-Linux OSes
1536 are unable to enable it back. This is intended to be used in environments
1537 where stopping a proxy is not even imaginable but a misconfigured proxy must
1538 be fixed. That way it's possible to release the port and bind it into another
1539 process to restore operations. The frontend will appear with status "STOP"
1540 on the stats page.
1541
1542 The frontend may be specified either by its name or by its numeric ID,
1543 prefixed with a sharp ('#').
1544
1545 This command is restricted and can only be issued on sockets configured for
1546 level "admin".
1547
1548disable health <backend>/<server>
1549 Mark the primary health check as temporarily stopped. This will disable
1550 sending of health checks, and the last health check result will be ignored.
1551 The server will be in unchecked state and considered UP unless an auxiliary
1552 agent check forces it down.
1553
1554 This command is restricted and can only be issued on sockets configured for
1555 level "admin".
1556
1557disable server <backend>/<server>
1558 Mark the server DOWN for maintenance. In this mode, no more checks will be
1559 performed on the server until it leaves maintenance.
1560 If the server is tracked by other servers, those servers will be set to DOWN
1561 during the maintenance.
1562
1563 In the statistics page, a server DOWN for maintenance will appear with a
1564 "MAINT" status, its tracking servers with the "MAINT(via)" one.
1565
1566 Both the backend and the server may be specified either by their name or by
1567 their numeric ID, prefixed with a sharp ('#').
1568
1569 This command is restricted and can only be issued on sockets configured for
1570 level "admin".
1571
1572enable agent <backend>/<server>
1573 Resume auxiliary agent check that was temporarily stopped.
1574
1575 See "disable agent" for details of the effect of temporarily starting
1576 and stopping an auxiliary agent.
1577
1578 This command is restricted and can only be issued on sockets configured for
1579 level "admin".
1580
Olivier Houchard614f8d72017-03-14 20:08:46 +01001581enable dynamic-cookie backend <backend>
n9@users.noreply.github.com25a1c8e2019-08-23 11:21:05 +02001582 Enable the generation of dynamic cookies for the backend <backend>.
1583 A secret key must also be provided.
Olivier Houchard614f8d72017-03-14 20:08:46 +01001584
Willy Tarreau44aed902015-10-13 14:45:29 +02001585enable frontend <frontend>
1586 Resume a frontend which was temporarily stopped. It is possible that some of
1587 the listening ports won't be able to bind anymore (eg: if another process
1588 took them since the 'disable frontend' operation). If this happens, an error
1589 is displayed. Some operating systems might not be able to resume a frontend
1590 which was disabled.
1591
1592 The frontend may be specified either by its name or by its numeric ID,
1593 prefixed with a sharp ('#').
1594
1595 This command is restricted and can only be issued on sockets configured for
1596 level "admin".
1597
1598enable health <backend>/<server>
1599 Resume a primary health check that was temporarily stopped. This will enable
1600 sending of health checks again. Please see "disable health" for details.
1601
1602 This command is restricted and can only be issued on sockets configured for
1603 level "admin".
1604
1605enable server <backend>/<server>
1606 If the server was previously marked as DOWN for maintenance, this marks the
1607 server UP and checks are re-enabled.
1608
1609 Both the backend and the server may be specified either by their name or by
1610 their numeric ID, prefixed with a sharp ('#').
1611
1612 This command is restricted and can only be issued on sockets configured for
1613 level "admin".
1614
Willy Tarreauabb9f9b2019-10-24 17:55:53 +02001615expert-mode [on|off]
1616 Without options, this indicates whether the expert mode is enabled or
1617 disabled on the current connection. When passed "on", it turns the expert
1618 mode on for the current CLI connection only. With "off" it turns it off. The
1619 expert mode enables displaying of expert commands that can be extremely
1620 dangerous for the process and which may occasionally help developers collect
1621 important information about complex bugs. Any misuse of these features will
1622 likely lead to a process crash. Do not use this option without being invited
1623 to do so. Note that this command is purposely not listed in the help message.
1624 This command is only accessible in admin level. Changing to another level
1625 automatically resets the expert mode.
1626
Willy Tarreau44aed902015-10-13 14:45:29 +02001627get map <map> <value>
1628get acl <acl> <value>
1629 Lookup the value <value> in the map <map> or in the ACL <acl>. <map> or <acl>
1630 are the #<id> or the <file> returned by "show map" or "show acl". This command
1631 returns all the matching patterns associated with this map. This is useful for
1632 debugging maps and ACLs. The output format is composed by one line par
1633 matching type. Each line is composed by space-delimited series of words.
1634
1635 The first two words are:
1636
1637 <match method>: The match method applied. It can be "found", "bool",
1638 "int", "ip", "bin", "len", "str", "beg", "sub", "dir",
1639 "dom", "end" or "reg".
1640
1641 <match result>: The result. Can be "match" or "no-match".
1642
1643 The following words are returned only if the pattern matches an entry.
1644
1645 <index type>: "tree" or "list". The internal lookup algorithm.
1646
1647 <case>: "case-insensitive" or "case-sensitive". The
1648 interpretation of the case.
1649
1650 <entry matched>: match="<entry>". Return the matched pattern. It is
1651 useful with regular expressions.
1652
1653 The two last word are used to show the returned value and its type. With the
1654 "acl" case, the pattern doesn't exist.
1655
1656 return=nothing: No return because there are no "map".
1657 return="<value>": The value returned in the string format.
1658 return=cannot-display: The value cannot be converted as string.
1659
1660 type="<type>": The type of the returned sample.
1661
1662get weight <backend>/<server>
1663 Report the current weight and the initial weight of server <server> in
1664 backend <backend> or an error if either doesn't exist. The initial weight is
1665 the one that appears in the configuration file. Both are normally equal
1666 unless the current weight has been changed. Both the backend and the server
1667 may be specified either by their name or by their numeric ID, prefixed with a
1668 sharp ('#').
1669
1670help
1671 Print the list of known keywords and their basic usage. The same help screen
1672 is also displayed for unknown commands.
1673
1674prompt
1675 Toggle the prompt at the beginning of the line and enter or leave interactive
1676 mode. In interactive mode, the connection is not closed after a command
1677 completes. Instead, the prompt will appear again, indicating the user that
1678 the interpreter is waiting for a new command. The prompt consists in a right
1679 angle bracket followed by a space "> ". This mode is particularly convenient
1680 when one wants to periodically check information such as stats or errors.
1681 It is also a good idea to enter interactive mode before issuing a "help"
1682 command.
1683
1684quit
1685 Close the connection when in interactive mode.
1686
Olivier Houchard614f8d72017-03-14 20:08:46 +01001687set dynamic-cookie-key backend <backend> <value>
1688 Modify the secret key used to generate the dynamic persistent cookies.
1689 This will break the existing sessions.
1690
Willy Tarreau44aed902015-10-13 14:45:29 +02001691set map <map> [<key>|#<ref>] <value>
1692 Modify the value corresponding to each key <key> in a map <map>. <map> is the
1693 #<id> or <file> returned by "show map". If the <ref> is used in place of
1694 <key>, only the entry pointed by <ref> is changed. The new value is <value>.
1695
1696set maxconn frontend <frontend> <value>
1697 Dynamically change the specified frontend's maxconn setting. Any positive
1698 value is allowed including zero, but setting values larger than the global
1699 maxconn does not make much sense. If the limit is increased and connections
1700 were pending, they will immediately be accepted. If it is lowered to a value
1701 below the current number of connections, new connections acceptation will be
1702 delayed until the threshold is reached. The frontend might be specified by
1703 either its name or its numeric ID prefixed with a sharp ('#').
1704
Andrew Hayworthedb93a72015-10-27 21:46:25 +00001705set maxconn server <backend/server> <value>
1706 Dynamically change the specified server's maxconn setting. Any positive
1707 value is allowed including zero, but setting values larger than the global
1708 maxconn does not make much sense.
1709
Willy Tarreau44aed902015-10-13 14:45:29 +02001710set maxconn global <maxconn>
1711 Dynamically change the global maxconn setting within the range defined by the
1712 initial global maxconn setting. If it is increased and connections were
1713 pending, they will immediately be accepted. If it is lowered to a value below
1714 the current number of connections, new connections acceptation will be
1715 delayed until the threshold is reached. A value of zero restores the initial
1716 setting.
1717
Willy Tarreaud2d33482019-04-25 17:09:07 +02001718set profiling { tasks } { auto | on | off }
Willy Tarreau75c62c22018-11-22 11:02:09 +01001719 Enables or disables CPU profiling for the indicated subsystem. This is
1720 equivalent to setting or clearing the "profiling" settings in the "global"
1721 section of the configuration file. Please also see "show profiling".
1722
Willy Tarreau44aed902015-10-13 14:45:29 +02001723set rate-limit connections global <value>
1724 Change the process-wide connection rate limit, which is set by the global
1725 'maxconnrate' setting. A value of zero disables the limitation. This limit
1726 applies to all frontends and the change has an immediate effect. The value
1727 is passed in number of connections per second.
1728
1729set rate-limit http-compression global <value>
1730 Change the maximum input compression rate, which is set by the global
1731 'maxcomprate' setting. A value of zero disables the limitation. The value is
1732 passed in number of kilobytes per second. The value is available in the "show
1733 info" on the line "CompressBpsRateLim" in bytes.
1734
1735set rate-limit sessions global <value>
1736 Change the process-wide session rate limit, which is set by the global
1737 'maxsessrate' setting. A value of zero disables the limitation. This limit
1738 applies to all frontends and the change has an immediate effect. The value
1739 is passed in number of sessions per second.
1740
1741set rate-limit ssl-sessions global <value>
1742 Change the process-wide SSL session rate limit, which is set by the global
1743 'maxsslrate' setting. A value of zero disables the limitation. This limit
1744 applies to all frontends and the change has an immediate effect. The value
1745 is passed in number of sessions per second sent to the SSL stack. It applies
1746 before the handshake in order to protect the stack against handshake abuses.
1747
Baptiste Assmann3749ebf2016-08-03 22:34:12 +02001748set server <backend>/<server> addr <ip4 or ip6 address> [port <port>]
Willy Tarreau44aed902015-10-13 14:45:29 +02001749 Replace the current IP address of a server by the one provided.
Michael Prokop4438c602019-05-24 10:25:45 +02001750 Optionally, the port can be changed using the 'port' parameter.
Baptiste Assmann3749ebf2016-08-03 22:34:12 +02001751 Note that changing the port also support switching from/to port mapping
1752 (notation with +X or -Y), only if a port is configured for the health check.
Willy Tarreau44aed902015-10-13 14:45:29 +02001753
1754set server <backend>/<server> agent [ up | down ]
1755 Force a server's agent to a new state. This can be useful to immediately
1756 switch a server's state regardless of some slow agent checks for example.
1757 Note that the change is propagated to tracking servers if any.
1758
Misiek43972902017-01-09 09:53:06 +01001759set server <backend>/<server> agent-addr <addr>
1760 Change addr for servers agent checks. Allows to migrate agent-checks to
1761 another address at runtime. You can specify both IP and hostname, it will be
1762 resolved.
1763
1764set server <backend>/<server> agent-send <value>
1765 Change agent string sent to agent check target. Allows to update string while
1766 changing server address to keep those two matching.
1767
Willy Tarreau44aed902015-10-13 14:45:29 +02001768set server <backend>/<server> health [ up | stopping | down ]
1769 Force a server's health to a new state. This can be useful to immediately
1770 switch a server's state regardless of some slow health checks for example.
1771 Note that the change is propagated to tracking servers if any.
1772
Baptiste Assmann50946562016-08-31 23:26:29 +02001773set server <backend>/<server> check-port <port>
1774 Change the port used for health checking to <port>
1775
Willy Tarreau44aed902015-10-13 14:45:29 +02001776set server <backend>/<server> state [ ready | drain | maint ]
1777 Force a server's administrative state to a new state. This can be useful to
1778 disable load balancing and/or any traffic to a server. Setting the state to
1779 "ready" puts the server in normal mode, and the command is the equivalent of
1780 the "enable server" command. Setting the state to "maint" disables any traffic
1781 to the server as well as any health checks. This is the equivalent of the
1782 "disable server" command. Setting the mode to "drain" only removes the server
1783 from load balancing but still allows it to be checked and to accept new
1784 persistent connections. Changes are propagated to tracking servers if any.
1785
1786set server <backend>/<server> weight <weight>[%]
1787 Change a server's weight to the value passed in argument. This is the exact
1788 equivalent of the "set weight" command below.
1789
Frédéric Lécailleb418c122017-04-26 11:24:02 +02001790set server <backend>/<server> fqdn <FQDN>
Lukas Tribusc5dd5a52018-08-14 11:39:35 +02001791 Change a server's FQDN to the value passed in argument. This requires the
1792 internal run-time DNS resolver to be configured and enabled for this server.
Frédéric Lécailleb418c122017-04-26 11:24:02 +02001793
Andjelko Iharosc4df59e2017-07-20 11:59:48 +02001794set severity-output [ none | number | string ]
1795 Change the severity output format of the stats socket connected to for the
1796 duration of the current session.
1797
William Lallemand6ab08b32019-11-29 16:48:43 +01001798set ssl cert <filename> <payload>
1799 This command is part of a transaction system, the "commit ssl cert" and
1800 "abort ssl cert" commands could be required.
1801 If there is no on-going transaction, it will duplicate the certificate
1802 <filename> in memory to a temporary transaction, then update this
1803 transaction with the PEM file in the payload. If a transaction exists with
1804 the same filename, it will update this transaction. It's also possible to
1805 update the files linked to a certificate (.issuer, .sctl, .oscp etc.)
1806 Once the modification are done, you have to "commit ssl cert" the
1807 transaction.
1808
1809 Example:
1810 echo -e "set ssl cert localhost.pem <<\n$(cat 127.0.0.1.pem)\n" | \
1811 socat /var/run/haproxy.stat -
1812 echo -e \
1813 "set ssl cert localhost.pem.issuer <<\n $(cat 127.0.0.1.pem.issuer)\n" | \
1814 socat /var/run/haproxy.stat -
1815 echo -e \
1816 "set ssl cert localhost.pem.ocsp <<\n$(base64 -w 1000 127.0.0.1.pem.ocsp)\n" | \
1817 socat /var/run/haproxy.stat -
1818 echo "commit ssl cert localhost.pem" | socat /var/run/haproxy.stat -
1819
Aurélien Nephtali1e0867c2018-04-18 14:04:58 +02001820set ssl ocsp-response <response | payload>
Willy Tarreau44aed902015-10-13 14:45:29 +02001821 This command is used to update an OCSP Response for a certificate (see "crt"
1822 on "bind" lines). Same controls are performed as during the initial loading of
1823 the response. The <response> must be passed as a base64 encoded string of the
Emmanuel Hocdet2c32d8f2017-05-22 14:58:00 +02001824 DER encoded response from the OCSP server. This command is not supported with
1825 BoringSSL.
Willy Tarreau44aed902015-10-13 14:45:29 +02001826
1827 Example:
1828 openssl ocsp -issuer issuer.pem -cert server.pem \
1829 -host ocsp.issuer.com:80 -respout resp.der
1830 echo "set ssl ocsp-response $(base64 -w 10000 resp.der)" | \
1831 socat stdio /var/run/haproxy.stat
1832
Aurélien Nephtali1e0867c2018-04-18 14:04:58 +02001833 using the payload syntax:
1834 echo -e "set ssl ocsp-response <<\n$(base64 resp.der)\n" | \
1835 socat stdio /var/run/haproxy.stat
1836
Willy Tarreau44aed902015-10-13 14:45:29 +02001837set ssl tls-key <id> <tlskey>
1838 Set the next TLS key for the <id> listener to <tlskey>. This key becomes the
1839 ultimate key, while the penultimate one is used for encryption (others just
1840 decrypt). The oldest TLS key present is overwritten. <id> is either a numeric
1841 #<id> or <file> returned by "show tls-keys". <tlskey> is a base64 encoded 48
Emeric Brun9e754772019-01-10 17:51:55 +01001842 or 80 bits TLS ticket key (ex. openssl rand 80 | openssl base64 -A).
Willy Tarreau44aed902015-10-13 14:45:29 +02001843
1844set table <table> key <key> [data.<data_type> <value>]*
1845 Create or update a stick-table entry in the table. If the key is not present,
1846 an entry is inserted. See stick-table in section 4.2 to find all possible
1847 values for <data_type>. The most likely use consists in dynamically entering
1848 entries for source IP addresses, with a flag in gpc0 to dynamically block an
1849 IP address or affect its quality of service. It is possible to pass multiple
1850 data_types in a single call.
1851
1852set timeout cli <delay>
1853 Change the CLI interface timeout for current connection. This can be useful
1854 during long debugging sessions where the user needs to constantly inspect
1855 some indicators without being disconnected. The delay is passed in seconds.
1856
1857set weight <backend>/<server> <weight>[%]
1858 Change a server's weight to the value passed in argument. If the value ends
1859 with the '%' sign, then the new weight will be relative to the initially
1860 configured weight. Absolute weights are permitted between 0 and 256.
1861 Relative weights must be positive with the resulting absolute weight is
1862 capped at 256. Servers which are part of a farm running a static
1863 load-balancing algorithm have stricter limitations because the weight
1864 cannot change once set. Thus for these servers, the only accepted values
1865 are 0 and 100% (or 0 and the initial weight). Changes take effect
1866 immediately, though certain LB algorithms require a certain amount of
1867 requests to consider changes. A typical usage of this command is to
1868 disable a server during an update by setting its weight to zero, then to
1869 enable it again after the update by setting it back to 100%. This command
1870 is restricted and can only be issued on sockets configured for level
1871 "admin". Both the backend and the server may be specified either by their
1872 name or by their numeric ID, prefixed with a sharp ('#').
1873
Willy Tarreaud6129fc2017-07-28 16:52:23 +02001874show acl [<acl>]
1875 Dump info about acl converters. Without argument, the list of all available
1876 acls is returned. If a <acl> is specified, its contents are dumped. <acl> if
1877 the #<id> or <file>. The dump format is the same than the map even for the
1878 sample value. The data returned are not a list of available ACL, but are the
1879 list of all patterns composing any ACL. Many of these patterns can be shared
1880 with maps.
1881
1882show backend
1883 Dump the list of backends available in the running process
1884
William Lallemand67a234f2018-12-13 09:05:45 +01001885show cli level
1886 Display the CLI level of the current CLI session. The result could be
1887 'admin', 'operator' or 'user'. See also the 'operator' and 'user' commands.
1888
1889 Example :
1890
1891 $ socat /tmp/sock1 readline
1892 prompt
1893 > operator
1894 > show cli level
1895 operator
1896 > user
1897 > show cli level
1898 user
1899 > operator
1900 Permission denied
1901
1902operator
1903 Decrease the CLI level of the current CLI session to operator. It can't be
Willy Tarreauabb9f9b2019-10-24 17:55:53 +02001904 increased. It also drops expert mode. See also "show cli level".
William Lallemand67a234f2018-12-13 09:05:45 +01001905
1906user
1907 Decrease the CLI level of the current CLI session to user. It can't be
Willy Tarreauabb9f9b2019-10-24 17:55:53 +02001908 increased. It also drops expert mode. See also "show cli level".
William Lallemand67a234f2018-12-13 09:05:45 +01001909
Willy Tarreau4c356932019-05-16 17:39:32 +02001910show activity
1911 Reports some counters about internal events that will help developers and
1912 more generally people who know haproxy well enough to narrow down the causes
1913 of reports of abnormal behaviours. A typical example would be a properly
1914 running process never sleeping and eating 100% of the CPU. The output fields
1915 will be made of one line per metric, and per-thread counters on the same
1916 line. These counters are 32-bit and will wrap during the process' life, which
1917 is not a problem since calls to this command will typically be performed
1918 twice. The fields are purposely not documented so that their exact meaning is
1919 verified in the code where the counters are fed. These values are also reset
1920 by the "clear counters" command.
1921
William Lallemand51132162016-12-16 16:38:58 +01001922show cli sockets
1923 List CLI sockets. The output format is composed of 3 fields separated by
1924 spaces. The first field is the socket address, it can be a unix socket, a
1925 ipv4 address:port couple or a ipv6 one. Socket of other types won't be dump.
1926 The second field describe the level of the socket: 'admin', 'user' or
1927 'operator'. The last field list the processes on which the socket is bound,
1928 separated by commas, it can be numbers or 'all'.
1929
1930 Example :
1931
1932 $ echo 'show cli sockets' | socat stdio /tmp/sock1
1933 # socket lvl processes
1934 /tmp/sock1 admin all
1935 127.0.0.1:9999 user 2,3,4
1936 127.0.0.2:9969 user 2
1937 [::1]:9999 operator 2
1938
William Lallemand86d0df02017-11-24 21:36:45 +01001939show cache
Cyril Bonté7b888f12017-11-26 22:24:31 +01001940 List the configured caches and the objects stored in each cache tree.
William Lallemand86d0df02017-11-24 21:36:45 +01001941
1942 $ echo 'show cache' | socat stdio /tmp/sock1
1943 0x7f6ac6c5b03a: foobar (shctx:0x7f6ac6c5b000, available blocks:3918)
1944 1 2 3 4
1945
1946 1. pointer to the cache structure
1947 2. cache name
1948 3. pointer to the mmap area (shctx)
1949 4. number of blocks available for reuse in the shctx
1950
1951 0x7f6ac6c5b4cc hash:286881868 size:39114 (39 blocks), refcount:9, expire:237
1952 1 2 3 4 5 6
1953
1954 1. pointer to the cache entry
1955 2. first 32 bits of the hash
1956 3. size of the object in bytes
1957 4. number of blocks used for the object
1958 5. number of transactions using the entry
1959 6. expiration time, can be negative if already expired
1960
Willy Tarreauae795722016-02-16 11:27:28 +01001961show env [<name>]
1962 Dump one or all environment variables known by the process. Without any
1963 argument, all variables are dumped. With an argument, only the specified
1964 variable is dumped if it exists. Otherwise "Variable not found" is emitted.
1965 Variables are dumped in the same format as they are stored or returned by the
1966 "env" utility, that is, "<name>=<value>". This can be handy when debugging
1967 certain configuration files making heavy use of environment variables to
1968 ensure that they contain the expected values. This command is restricted and
1969 can only be issued on sockets configured for levels "operator" or "admin".
1970
Willy Tarreau35069f82016-11-25 09:16:37 +01001971show errors [<iid>|<proxy>] [request|response]
Willy Tarreau44aed902015-10-13 14:45:29 +02001972 Dump last known request and response errors collected by frontends and
1973 backends. If <iid> is specified, the limit the dump to errors concerning
Willy Tarreau234ba2d2016-11-25 08:39:10 +01001974 either frontend or backend whose ID is <iid>. Proxy ID "-1" will cause
1975 all instances to be dumped. If a proxy name is specified instead, its ID
Willy Tarreau35069f82016-11-25 09:16:37 +01001976 will be used as the filter. If "request" or "response" is added after the
1977 proxy name or ID, only request or response errors will be dumped. This
1978 command is restricted and can only be issued on sockets configured for
1979 levels "operator" or "admin".
Willy Tarreau44aed902015-10-13 14:45:29 +02001980
1981 The errors which may be collected are the last request and response errors
1982 caused by protocol violations, often due to invalid characters in header
1983 names. The report precisely indicates what exact character violated the
1984 protocol. Other important information such as the exact date the error was
1985 detected, frontend and backend names, the server name (when known), the
1986 internal session ID and the source address which has initiated the session
1987 are reported too.
1988
1989 All characters are returned, and non-printable characters are encoded. The
1990 most common ones (\t = 9, \n = 10, \r = 13 and \e = 27) are encoded as one
1991 letter following a backslash. The backslash itself is encoded as '\\' to
1992 avoid confusion. Other non-printable characters are encoded '\xNN' where
1993 NN is the two-digits hexadecimal representation of the character's ASCII
1994 code.
1995
1996 Lines are prefixed with the position of their first character, starting at 0
1997 for the beginning of the buffer. At most one input line is printed per line,
1998 and large lines will be broken into multiple consecutive output lines so that
1999 the output never goes beyond 79 characters wide. It is easy to detect if a
2000 line was broken, because it will not end with '\n' and the next line's offset
2001 will be followed by a '+' sign, indicating it is a continuation of previous
2002 line.
2003
2004 Example :
Willy Tarreau35069f82016-11-25 09:16:37 +01002005 $ echo "show errors -1 response" | socat stdio /tmp/sock1
Willy Tarreau44aed902015-10-13 14:45:29 +02002006 >>> [04/Mar/2009:15:46:56.081] backend http-in (#2) : invalid response
2007 src 127.0.0.1, session #54, frontend fe-eth0 (#1), server s2 (#1)
2008 response length 213 bytes, error at position 23:
2009
2010 00000 HTTP/1.0 200 OK\r\n
2011 00017 header/bizarre:blah\r\n
2012 00038 Location: blah\r\n
2013 00054 Long-line: this is a very long line which should b
2014 00104+ e broken into multiple lines on the output buffer,
2015 00154+ otherwise it would be too large to print in a ter
2016 00204+ minal\r\n
2017 00211 \r\n
2018
2019 In the example above, we see that the backend "http-in" which has internal
2020 ID 2 has blocked an invalid response from its server s2 which has internal
2021 ID 1. The request was on session 54 initiated by source 127.0.0.1 and
2022 received by frontend fe-eth0 whose ID is 1. The total response length was
2023 213 bytes when the error was detected, and the error was at byte 23. This
2024 is the slash ('/') in header name "header/bizarre", which is not a valid
2025 HTTP character for a header name.
2026
Willy Tarreau1d181e42019-08-30 11:17:01 +02002027show events [<sink>] [-w] [-n]
Willy Tarreau9f830d72019-08-26 18:17:04 +02002028 With no option, this lists all known event sinks and their types. With an
2029 option, it will dump all available events in the designated sink if it is of
Willy Tarreau1d181e42019-08-30 11:17:01 +02002030 type buffer. If option "-w" is passed after the sink name, then once the end
2031 of the buffer is reached, the command will wait for new events and display
2032 them. It is possible to stop the operation by entering any input (which will
2033 be discarded) or by closing the session. Finally, option "-n" is used to
2034 directly seek to the end of the buffer, which is often convenient when
2035 combined with "-w" to only report new events. For convenience, "-wn" or "-nw"
2036 may be used to enable both options at once.
Willy Tarreau9f830d72019-08-26 18:17:04 +02002037
Willy Tarreau7a4a0ac2017-07-25 19:32:50 +02002038show fd [<fd>]
2039 Dump the list of either all open file descriptors or just the one number <fd>
2040 if specified. This is only aimed at developers who need to observe internal
2041 states in order to debug complex issues such as abnormal CPU usages. One fd
2042 is reported per lines, and for each of them, its state in the poller using
2043 upper case letters for enabled flags and lower case for disabled flags, using
2044 "P" for "polled", "R" for "ready", "A" for "active", the events status using
2045 "H" for "hangup", "E" for "error", "O" for "output", "P" for "priority" and
2046 "I" for "input", a few other flags like "N" for "new" (just added into the fd
2047 cache), "U" for "updated" (received an update in the fd cache), "L" for
2048 "linger_risk", "C" for "cloned", then the cached entry position, the pointer
2049 to the internal owner, the pointer to the I/O callback and its name when
2050 known. When the owner is a connection, the connection flags, and the target
2051 are reported (frontend, proxy or server). When the owner is a listener, the
2052 listener's state and its frontend are reported. There is no point in using
2053 this command without a good knowledge of the internals. It's worth noting
2054 that the output format may evolve over time so this output must not be parsed
2055 by tools designed to be durable.
2056
Willy Tarreau6b19b142019-10-09 15:44:21 +02002057show info [typed|json] [desc]
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002058 Dump info about haproxy status on current process. If "typed" is passed as an
2059 optional argument, field numbers, names and types are emitted as well so that
2060 external monitoring products can easily retrieve, possibly aggregate, then
2061 report information found in fields they don't know. Each field is dumped on
Simon Horman05ee2132017-01-04 09:37:25 +01002062 its own line. If "json" is passed as an optional argument then
2063 information provided by "typed" output is provided in JSON format as a
2064 list of JSON objects. By default, the format contains only two columns
2065 delimited by a colon (':'). The left one is the field name and the right
2066 one is the value. It is very important to note that in typed output
2067 format, the dump for a single object is contiguous so that there is no
2068 need for a consumer to store everything at once.
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002069
2070 When using the typed output format, each line is made of 4 columns delimited
2071 by colons (':'). The first column is a dot-delimited series of 3 elements. The
2072 first element is the numeric position of the field in the list (starting at
2073 zero). This position shall not change over time, but holes are to be expected,
2074 depending on build options or if some fields are deleted in the future. The
2075 second element is the field name as it appears in the default "show info"
2076 output. The third element is the relative process number starting at 1.
2077
2078 The rest of the line starting after the first colon follows the "typed output
2079 format" described in the section above. In short, the second column (after the
2080 first ':') indicates the origin, nature and scope of the variable. The third
2081 column indicates the type of the field, among "s32", "s64", "u32", "u64" and
2082 "str". Then the fourth column is the value itself, which the consumer knows
2083 how to parse thanks to column 3 and how to process thanks to column 2.
2084
2085 Thus the overall line format in typed mode is :
2086
2087 <field_pos>.<field_name>.<process_num>:<tags>:<type>:<value>
2088
Willy Tarreau6b19b142019-10-09 15:44:21 +02002089 When "desc" is appended to the command, one extra colon followed by a quoted
2090 string is appended with a description for the metric. At the time of writing,
2091 this is only supported for the "typed" and default output formats.
2092
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002093 Example :
2094
2095 > show info
2096 Name: HAProxy
2097 Version: 1.7-dev1-de52ea-146
2098 Release_date: 2016/03/11
2099 Nbproc: 1
2100 Process_num: 1
2101 Pid: 28105
2102 Uptime: 0d 0h00m04s
2103 Uptime_sec: 4
2104 Memmax_MB: 0
2105 PoolAlloc_MB: 0
2106 PoolUsed_MB: 0
2107 PoolFailed: 0
2108 (...)
2109
2110 > show info typed
2111 0.Name.1:POS:str:HAProxy
2112 1.Version.1:POS:str:1.7-dev1-de52ea-146
2113 2.Release_date.1:POS:str:2016/03/11
2114 3.Nbproc.1:CGS:u32:1
2115 4.Process_num.1:KGP:u32:1
2116 5.Pid.1:SGP:u32:28105
2117 6.Uptime.1:MDP:str:0d 0h00m08s
2118 7.Uptime_sec.1:MDP:u32:8
2119 8.Memmax_MB.1:CLP:u32:0
2120 9.PoolAlloc_MB.1:MGP:u32:0
2121 10.PoolUsed_MB.1:MGP:u32:0
2122 11.PoolFailed.1:MCP:u32:0
2123 (...)
2124
Simon Horman1084a362016-11-21 17:00:24 +01002125 In the typed format, the presence of the process ID at the end of the
2126 first column makes it very easy to visually aggregate outputs from
2127 multiple processes.
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002128 Example :
2129
2130 $ ( echo show info typed | socat /var/run/haproxy.sock1 ; \
2131 echo show info typed | socat /var/run/haproxy.sock2 ) | \
2132 sort -t . -k 1,1n -k 2,2 -k 3,3n
2133 0.Name.1:POS:str:HAProxy
2134 0.Name.2:POS:str:HAProxy
2135 1.Version.1:POS:str:1.7-dev1-868ab3-148
2136 1.Version.2:POS:str:1.7-dev1-868ab3-148
2137 2.Release_date.1:POS:str:2016/03/11
2138 2.Release_date.2:POS:str:2016/03/11
2139 3.Nbproc.1:CGS:u32:2
2140 3.Nbproc.2:CGS:u32:2
2141 4.Process_num.1:KGP:u32:1
2142 4.Process_num.2:KGP:u32:2
2143 5.Pid.1:SGP:u32:30120
2144 5.Pid.2:SGP:u32:30121
2145 6.Uptime.1:MDP:str:0d 0h01m28s
2146 6.Uptime.2:MDP:str:0d 0h01m28s
2147 (...)
Willy Tarreau44aed902015-10-13 14:45:29 +02002148
Simon Horman05ee2132017-01-04 09:37:25 +01002149 The format of JSON output is described in a schema which may be output
Simon Horman6f6bb382017-01-04 09:37:26 +01002150 using "show schema json".
Simon Horman05ee2132017-01-04 09:37:25 +01002151
2152 The JSON output contains no extra whitespace in order to reduce the
2153 volume of output. For human consumption passing the output through a
2154 pretty printer may be helpful. Example :
2155
2156 $ echo "show info json" | socat /var/run/haproxy.sock stdio | \
2157 python -m json.tool
2158
Simon Horman6f6bb382017-01-04 09:37:26 +01002159 The JSON output contains no extra whitespace in order to reduce the
2160 volume of output. For human consumption passing the output through a
2161 pretty printer may be helpful. Example :
2162
2163 $ echo "show info json" | socat /var/run/haproxy.sock stdio | \
2164 python -m json.tool
2165
Willy Tarreau44aed902015-10-13 14:45:29 +02002166show map [<map>]
2167 Dump info about map converters. Without argument, the list of all available
2168 maps is returned. If a <map> is specified, its contents are dumped. <map> is
2169 the #<id> or <file>. The first column is a unique identifier. It can be used
2170 as reference for the operation "del map" and "set map". The second column is
2171 the pattern and the third column is the sample if available. The data returned
2172 are not directly a list of available maps, but are the list of all patterns
2173 composing any map. Many of these patterns can be shared with ACL.
2174
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002175show peers [<peers section>]
2176 Dump info about the peers configured in "peers" sections. Without argument,
2177 the list of the peers belonging to all the "peers" sections are listed. If
2178 <peers section> is specified, only the information about the peers belonging
2179 to this "peers" section are dumped.
2180
Michael Prokop4438c602019-05-24 10:25:45 +02002181 Here are two examples of outputs where hostA, hostB and hostC peers belong to
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002182 "sharedlb" peers sections. Only hostA and hostB are connected. Only hostA has
2183 sent data to hostB.
2184
2185 $ echo "show peers" | socat - /tmp/hostA
2186 0x55deb0224320: [15/Apr/2019:11:28:01] id=sharedlb state=0 flags=0x3 \
Emeric Brun0bbec0f2019-04-18 11:39:43 +02002187 resync_timeout=<PAST> task_calls=45122
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002188 0x55deb022b540: id=hostC(remote) addr=127.0.0.12:10002 status=CONN \
2189 reconnect=4s confirm=0
2190 flags=0x0
2191 0x55deb022a440: id=hostA(local) addr=127.0.0.10:10000 status=NONE \
2192 reconnect=<NEVER> confirm=0
2193 flags=0x0
2194 0x55deb0227d70: id=hostB(remote) addr=127.0.0.11:10001 status=ESTA
2195 reconnect=2s confirm=0
Emeric Brun0bbec0f2019-04-18 11:39:43 +02002196 flags=0x20000200 appctx:0x55deb028fba0 st0=7 st1=0 task_calls=14456 \
2197 state=EST
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002198 xprt=RAW src=127.0.0.1:37257 addr=127.0.0.10:10000
2199 remote_table:0x55deb0224a10 id=stkt local_id=1 remote_id=1
2200 last_local_table:0x55deb0224a10 id=stkt local_id=1 remote_id=1
2201 shared tables:
2202 0x55deb0224a10 local_id=1 remote_id=1 flags=0x0 remote_data=0x65
2203 last_acked=0 last_pushed=3 last_get=0 teaching_origin=0 update=3
2204 table:0x55deb022d6a0 id=stkt update=3 localupdate=3 \
2205 commitupdate=3 syncing=0
2206
2207 $ echo "show peers" | socat - /tmp/hostB
2208 0x55871b5ab320: [15/Apr/2019:11:28:03] id=sharedlb state=0 flags=0x3 \
Emeric Brun0bbec0f2019-04-18 11:39:43 +02002209 resync_timeout=<PAST> task_calls=3
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002210 0x55871b5b2540: id=hostC(remote) addr=127.0.0.12:10002 status=CONN \
2211 reconnect=3s confirm=0
2212 flags=0x0
2213 0x55871b5b1440: id=hostB(local) addr=127.0.0.11:10001 status=NONE \
2214 reconnect=<NEVER> confirm=0
2215 flags=0x0
2216 0x55871b5aed70: id=hostA(remote) addr=127.0.0.10:10000 status=ESTA \
2217 reconnect=2s confirm=0
Emeric Brun0bbec0f2019-04-18 11:39:43 +02002218 flags=0x20000200 appctx:0x7fa46800ee00 st0=7 st1=0 task_calls=62356 \
2219 state=EST
Frédéric Lécaille21dde502019-04-15 13:50:23 +02002220 remote_table:0x55871b5ab960 id=stkt local_id=1 remote_id=1
2221 last_local_table:0x55871b5ab960 id=stkt local_id=1 remote_id=1
2222 shared tables:
2223 0x55871b5ab960 local_id=1 remote_id=1 flags=0x0 remote_data=0x65
2224 last_acked=3 last_pushed=0 last_get=3 teaching_origin=0 update=0
2225 table:0x55871b5b46a0 id=stkt update=1 localupdate=0 \
2226 commitupdate=0 syncing=0
2227
Willy Tarreau44aed902015-10-13 14:45:29 +02002228show pools
2229 Dump the status of internal memory pools. This is useful to track memory
2230 usage when suspecting a memory leak for example. It does exactly the same
2231 as the SIGQUIT when running in foreground except that it does not flush
2232 the pools.
2233
Willy Tarreau75c62c22018-11-22 11:02:09 +01002234show profiling
2235 Dumps the current profiling settings, one per line, as well as the command
2236 needed to change them.
2237
Willy Tarreau44aed902015-10-13 14:45:29 +02002238show servers state [<backend>]
2239 Dump the state of the servers found in the running configuration. A backend
2240 name or identifier may be provided to limit the output to this backend only.
2241
2242 The dump has the following format:
2243 - first line contains the format version (1 in this specification);
2244 - second line contains the column headers, prefixed by a sharp ('#');
2245 - third line and next ones contain data;
2246 - each line starting by a sharp ('#') is considered as a comment.
2247
Dan Lloyd8e48b872016-07-01 21:01:18 -04002248 Since multiple versions of the output may co-exist, below is the list of
Willy Tarreau44aed902015-10-13 14:45:29 +02002249 fields and their order per file format version :
2250 1:
2251 be_id: Backend unique id.
2252 be_name: Backend label.
2253 srv_id: Server unique id (in the backend).
2254 srv_name: Server label.
2255 srv_addr: Server IP address.
2256 srv_op_state: Server operational state (UP/DOWN/...).
Cyril Bonté5b2ce8a2016-11-02 00:19:58 +01002257 0 = SRV_ST_STOPPED
2258 The server is down.
2259 1 = SRV_ST_STARTING
2260 The server is warming up (up but
2261 throttled).
2262 2 = SRV_ST_RUNNING
2263 The server is fully up.
2264 3 = SRV_ST_STOPPING
2265 The server is up but soft-stopping
2266 (eg: 404).
Willy Tarreau44aed902015-10-13 14:45:29 +02002267 srv_admin_state: Server administrative state (MAINT/DRAIN/...).
Cyril Bonté5b2ce8a2016-11-02 00:19:58 +01002268 The state is actually a mask of values :
2269 0x01 = SRV_ADMF_FMAINT
2270 The server was explicitly forced into
2271 maintenance.
2272 0x02 = SRV_ADMF_IMAINT
2273 The server has inherited the maintenance
2274 status from a tracked server.
2275 0x04 = SRV_ADMF_CMAINT
2276 The server is in maintenance because of
2277 the configuration.
2278 0x08 = SRV_ADMF_FDRAIN
2279 The server was explicitly forced into
2280 drain state.
2281 0x10 = SRV_ADMF_IDRAIN
2282 The server has inherited the drain status
2283 from a tracked server.
Baptiste Assmann89aa7f32016-11-02 21:31:27 +01002284 0x20 = SRV_ADMF_RMAINT
2285 The server is in maintenance because of an
2286 IP address resolution failure.
Frédéric Lécailleb418c122017-04-26 11:24:02 +02002287 0x40 = SRV_ADMF_HMAINT
2288 The server FQDN was set from stats socket.
2289
Willy Tarreau44aed902015-10-13 14:45:29 +02002290 srv_uweight: User visible server's weight.
2291 srv_iweight: Server's initial weight.
2292 srv_time_since_last_change: Time since last operational change.
2293 srv_check_status: Last health check status.
2294 srv_check_result: Last check result (FAILED/PASSED/...).
Cyril Bonté5b2ce8a2016-11-02 00:19:58 +01002295 0 = CHK_RES_UNKNOWN
2296 Initialized to this by default.
2297 1 = CHK_RES_NEUTRAL
2298 Valid check but no status information.
2299 2 = CHK_RES_FAILED
2300 Check failed.
2301 3 = CHK_RES_PASSED
2302 Check succeeded and server is fully up
2303 again.
2304 4 = CHK_RES_CONDPASS
2305 Check reports the server doesn't want new
2306 sessions.
Willy Tarreau44aed902015-10-13 14:45:29 +02002307 srv_check_health: Checks rise / fall current counter.
2308 srv_check_state: State of the check (ENABLED/PAUSED/...).
Cyril Bonté5b2ce8a2016-11-02 00:19:58 +01002309 The state is actually a mask of values :
2310 0x01 = CHK_ST_INPROGRESS
2311 A check is currently running.
2312 0x02 = CHK_ST_CONFIGURED
2313 This check is configured and may be
2314 enabled.
2315 0x04 = CHK_ST_ENABLED
2316 This check is currently administratively
2317 enabled.
2318 0x08 = CHK_ST_PAUSED
2319 Checks are paused because of maintenance
2320 (health only).
Willy Tarreau44aed902015-10-13 14:45:29 +02002321 srv_agent_state: State of the agent check (ENABLED/PAUSED/...).
Cyril Bonté5b2ce8a2016-11-02 00:19:58 +01002322 This state uses the same mask values as
2323 "srv_check_state", adding this specific one :
2324 0x10 = CHK_ST_AGENT
2325 Check is an agent check (otherwise it's a
2326 health check).
Willy Tarreau44aed902015-10-13 14:45:29 +02002327 bk_f_forced_id: Flag to know if the backend ID is forced by
2328 configuration.
2329 srv_f_forced_id: Flag to know if the server's ID is forced by
2330 configuration.
Frédéric Lécailleb418c122017-04-26 11:24:02 +02002331 srv_fqdn: Server FQDN.
Frédéric Lécaille31694712017-08-01 08:47:19 +02002332 srv_port: Server port.
Baptiste Assmann6d0f38f2018-07-02 17:00:54 +02002333 srvrecord: DNS SRV record associated to this SRV.
Willy Tarreau44aed902015-10-13 14:45:29 +02002334
2335show sess
2336 Dump all known sessions. Avoid doing this on slow connections as this can
2337 be huge. This command is restricted and can only be issued on sockets
2338 configured for levels "operator" or "admin".
2339
2340show sess <id>
2341 Display a lot of internal information about the specified session identifier.
2342 This identifier is the first field at the beginning of the lines in the dumps
2343 of "show sess" (it corresponds to the session pointer). Those information are
2344 useless to most users but may be used by haproxy developers to troubleshoot a
2345 complex bug. The output format is intentionally not documented so that it can
2346 freely evolve depending on demands. You may find a description of all fields
2347 returned in src/dumpstats.c
2348
2349 The special id "all" dumps the states of all sessions, which must be avoided
2350 as much as possible as it is highly CPU intensive and can take a lot of time.
2351
Willy Tarreau6b19b142019-10-09 15:44:21 +02002352show stat [{<iid>|<proxy>} <type> <sid>] [typed|json] [desc]
Simon Horman05ee2132017-01-04 09:37:25 +01002353 Dump statistics using the CSV format; using the extended typed output
2354 format described in the section above if "typed" is passed after the
2355 other arguments; or in JSON if "json" is passed after the other arguments
2356 . By passing <id>, <type> and <sid>, it is possible to dump only selected
2357 items :
Willy Tarreaua1b1ed52016-11-25 08:50:58 +01002358 - <iid> is a proxy ID, -1 to dump everything. Alternatively, a proxy name
2359 <proxy> may be specified. In this case, this proxy's ID will be used as
2360 the ID selector.
Willy Tarreau44aed902015-10-13 14:45:29 +02002361 - <type> selects the type of dumpable objects : 1 for frontends, 2 for
2362 backends, 4 for servers, -1 for everything. These values can be ORed,
2363 for example:
2364 1 + 2 = 3 -> frontend + backend.
2365 1 + 2 + 4 = 7 -> frontend + backend + server.
2366 - <sid> is a server ID, -1 to dump everything from the selected proxy.
2367
2368 Example :
2369 $ echo "show info;show stat" | socat stdio unix-connect:/tmp/sock1
2370 >>> Name: HAProxy
2371 Version: 1.4-dev2-49
2372 Release_date: 2009/09/23
2373 Nbproc: 1
2374 Process_num: 1
2375 (...)
2376
2377 # pxname,svname,qcur,qmax,scur,smax,slim,stot,bin,bout,dreq, (...)
2378 stats,FRONTEND,,,0,0,1000,0,0,0,0,0,0,,,,,OPEN,,,,,,,,,1,1,0, (...)
2379 stats,BACKEND,0,0,0,0,1000,0,0,0,0,0,,0,0,0,0,UP,0,0,0,,0,250,(...)
2380 (...)
2381 www1,BACKEND,0,0,0,0,1000,0,0,0,0,0,,0,0,0,0,UP,1,1,0,,0,250, (...)
2382
2383 $
2384
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002385 In this example, two commands have been issued at once. That way it's easy to
2386 find which process the stats apply to in multi-process mode. This is not
2387 needed in the typed output format as the process number is reported on each
2388 line. Notice the empty line after the information output which marks the end
2389 of the first block. A similar empty line appears at the end of the second
2390 block (stats) so that the reader knows the output has not been truncated.
2391
2392 When "typed" is specified, the output format is more suitable to monitoring
2393 tools because it provides numeric positions and indicates the type of each
2394 output field. Each value stands on its own line with process number, element
2395 number, nature, origin and scope. This same format is available via the HTTP
2396 stats by passing ";typed" after the URI. It is very important to note that in
Dan Lloyd8e48b872016-07-01 21:01:18 -04002397 typed output format, the dump for a single object is contiguous so that there
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002398 is no need for a consumer to store everything at once.
2399
2400 When using the typed output format, each line is made of 4 columns delimited
2401 by colons (':'). The first column is a dot-delimited series of 5 elements. The
2402 first element is a letter indicating the type of the object being described.
2403 At the moment the following object types are known : 'F' for a frontend, 'B'
2404 for a backend, 'L' for a listener, and 'S' for a server. The second element
2405 The second element is a positive integer representing the unique identifier of
2406 the proxy the object belongs to. It is equivalent to the "iid" column of the
2407 CSV output and matches the value in front of the optional "id" directive found
2408 in the frontend or backend section. The third element is a positive integer
2409 containing the unique object identifier inside the proxy, and corresponds to
2410 the "sid" column of the CSV output. ID 0 is reported when dumping a frontend
2411 or a backend. For a listener or a server, this corresponds to their respective
2412 ID inside the proxy. The fourth element is the numeric position of the field
2413 in the list (starting at zero). This position shall not change over time, but
2414 holes are to be expected, depending on build options or if some fields are
2415 deleted in the future. The fifth element is the field name as it appears in
2416 the CSV output. The sixth element is a positive integer and is the relative
2417 process number starting at 1.
2418
2419 The rest of the line starting after the first colon follows the "typed output
2420 format" described in the section above. In short, the second column (after the
2421 first ':') indicates the origin, nature and scope of the variable. The third
2422 column indicates the type of the field, among "s32", "s64", "u32", "u64" and
2423 "str". Then the fourth column is the value itself, which the consumer knows
2424 how to parse thanks to column 3 and how to process thanks to column 2.
2425
Willy Tarreau6b19b142019-10-09 15:44:21 +02002426 When "desc" is appended to the command, one extra colon followed by a quoted
2427 string is appended with a description for the metric. At the time of writing,
2428 this is only supported for the "typed" output format.
2429
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002430 Thus the overall line format in typed mode is :
2431
2432 <obj>.<px_id>.<id>.<fpos>.<fname>.<process_num>:<tags>:<type>:<value>
2433
2434 Here's an example of typed output format :
2435
2436 $ echo "show stat typed" | socat stdio unix-connect:/tmp/sock1
2437 F.2.0.0.pxname.1:MGP:str:private-frontend
2438 F.2.0.1.svname.1:MGP:str:FRONTEND
2439 F.2.0.8.bin.1:MGP:u64:0
2440 F.2.0.9.bout.1:MGP:u64:0
2441 F.2.0.40.hrsp_2xx.1:MGP:u64:0
2442 L.2.1.0.pxname.1:MGP:str:private-frontend
2443 L.2.1.1.svname.1:MGP:str:sock-1
2444 L.2.1.17.status.1:MGP:str:OPEN
2445 L.2.1.73.addr.1:MGP:str:0.0.0.0:8001
2446 S.3.13.60.rtime.1:MCP:u32:0
2447 S.3.13.61.ttime.1:MCP:u32:0
2448 S.3.13.62.agent_status.1:MGP:str:L4TOUT
2449 S.3.13.64.agent_duration.1:MGP:u64:2001
2450 S.3.13.65.check_desc.1:MCP:str:Layer4 timeout
2451 S.3.13.66.agent_desc.1:MCP:str:Layer4 timeout
2452 S.3.13.67.check_rise.1:MCP:u32:2
2453 S.3.13.68.check_fall.1:MCP:u32:3
2454 S.3.13.69.check_health.1:SGP:u32:0
2455 S.3.13.70.agent_rise.1:MaP:u32:1
2456 S.3.13.71.agent_fall.1:SGP:u32:1
2457 S.3.13.72.agent_health.1:SGP:u32:1
2458 S.3.13.73.addr.1:MCP:str:1.255.255.255:8888
2459 S.3.13.75.mode.1:MAP:str:http
2460 B.3.0.0.pxname.1:MGP:str:private-backend
2461 B.3.0.1.svname.1:MGP:str:BACKEND
2462 B.3.0.2.qcur.1:MGP:u32:0
2463 B.3.0.3.qmax.1:MGP:u32:0
2464 B.3.0.4.scur.1:MGP:u32:0
2465 B.3.0.5.smax.1:MGP:u32:0
2466 B.3.0.6.slim.1:MGP:u32:1000
2467 B.3.0.55.lastsess.1:MMP:s32:-1
2468 (...)
2469
Simon Horman1084a362016-11-21 17:00:24 +01002470 In the typed format, the presence of the process ID at the end of the
2471 first column makes it very easy to visually aggregate outputs from
2472 multiple processes, as show in the example below where each line appears
2473 for each process :
Willy Tarreau5d8b9792016-03-11 11:09:34 +01002474
2475 $ ( echo show stat typed | socat /var/run/haproxy.sock1 - ; \
2476 echo show stat typed | socat /var/run/haproxy.sock2 - ) | \
2477 sort -t . -k 1,1 -k 2,2n -k 3,3n -k 4,4n -k 5,5 -k 6,6n
2478 B.3.0.0.pxname.1:MGP:str:private-backend
2479 B.3.0.0.pxname.2:MGP:str:private-backend
2480 B.3.0.1.svname.1:MGP:str:BACKEND
2481 B.3.0.1.svname.2:MGP:str:BACKEND
2482 B.3.0.2.qcur.1:MGP:u32:0
2483 B.3.0.2.qcur.2:MGP:u32:0
2484 B.3.0.3.qmax.1:MGP:u32:0
2485 B.3.0.3.qmax.2:MGP:u32:0
2486 B.3.0.4.scur.1:MGP:u32:0
2487 B.3.0.4.scur.2:MGP:u32:0
2488 B.3.0.5.smax.1:MGP:u32:0
2489 B.3.0.5.smax.2:MGP:u32:0
2490 B.3.0.6.slim.1:MGP:u32:1000
2491 B.3.0.6.slim.2:MGP:u32:1000
2492 (...)
Willy Tarreau44aed902015-10-13 14:45:29 +02002493
Simon Horman05ee2132017-01-04 09:37:25 +01002494 The format of JSON output is described in a schema which may be output
Simon Horman6f6bb382017-01-04 09:37:26 +01002495 using "show schema json".
2496
2497 The JSON output contains no extra whitespace in order to reduce the
2498 volume of output. For human consumption passing the output through a
2499 pretty printer may be helpful. Example :
2500
2501 $ echo "show stat json" | socat /var/run/haproxy.sock stdio | \
2502 python -m json.tool
Simon Horman05ee2132017-01-04 09:37:25 +01002503
2504 The JSON output contains no extra whitespace in order to reduce the
2505 volume of output. For human consumption passing the output through a
2506 pretty printer may be helpful. Example :
2507
2508 $ echo "show stat json" | socat /var/run/haproxy.sock stdio | \
2509 python -m json.tool
2510
William Lallemandd4f946c2019-12-05 10:26:40 +01002511show ssl cert [<filename>]
2512 Display the list of certicates used on frontends. If a filename is prefixed
2513 by an asterisk, it is a transaction which is not commited yet. If a
2514 filename is specified, it will show details about the certificate. This
2515 command can be useful to check if a certificate was well updated. You can
2516 also display details on a transaction by prefixing the filename by an
2517 asterisk.
2518
2519 Example :
2520
2521 $ echo "@1 show ssl cert" | socat /var/run/haproxy.master -
2522 # transaction
2523 *test.local.pem
2524 # filename
2525 test.local.pem
2526
2527 $ echo "@1 show ssl cert test.local.pem" | socat /var/run/haproxy.master -
2528 Filename: test.local.pem
2529 Serial: 03ECC19BA54B25E85ABA46EE561B9A10D26F
2530 notBefore: Sep 13 21:20:24 2019 GMT
2531 notAfter: Dec 12 21:20:24 2019 GMT
2532 Issuer: /C=US/O=Let's Encrypt/CN=Let's Encrypt Authority X3
2533 Subject: /CN=test.local
2534 Subject Alternative Name: DNS:test.local, DNS:imap.test.local
2535 Algorithm: RSA2048
2536 SHA1 FingerPrint: 417A11CAE25F607B24F638B4A8AEE51D1E211477
2537
2538 $ echo "@1 show ssl cert *test.local.pem" | socat /var/run/haproxy.master -
2539 Filename: *test.local.pem
2540 [...]
2541
Christopher Faulet78c43062019-09-27 10:45:47 +02002542show resolvers [<resolvers section id>]
Willy Tarreau44aed902015-10-13 14:45:29 +02002543 Dump statistics for the given resolvers section, or all resolvers sections
2544 if no section is supplied.
2545
2546 For each name server, the following counters are reported:
2547 sent: number of DNS requests sent to this server
2548 valid: number of DNS valid responses received from this server
2549 update: number of DNS responses used to update the server's IP address
2550 cname: number of CNAME responses
2551 cname_error: CNAME errors encountered with this server
2552 any_err: number of empty response (IE: server does not support ANY type)
2553 nx: non existent domain response received from this server
2554 timeout: how many time this server did not answer in time
2555 refused: number of requests refused by this server
2556 other: any other DNS errors
2557 invalid: invalid DNS response (from a protocol point of view)
2558 too_big: too big response
2559 outdated: number of response arrived too late (after an other name server)
2560
2561show table
2562 Dump general information on all known stick-tables. Their name is returned
2563 (the name of the proxy which holds them), their type (currently zero, always
2564 IP), their size in maximum possible number of entries, and the number of
2565 entries currently in use.
2566
2567 Example :
2568 $ echo "show table" | socat stdio /tmp/sock1
2569 >>> # table: front_pub, type: ip, size:204800, used:171454
2570 >>> # table: back_rdp, type: ip, size:204800, used:0
2571
Adis Nezirovic1a693fc2020-01-16 15:19:29 +01002572show table <name> [ data.<type> <operator> <value> [data.<type> ...]] | [ key <key> ]
Willy Tarreau44aed902015-10-13 14:45:29 +02002573 Dump contents of stick-table <name>. In this mode, a first line of generic
2574 information about the table is reported as with "show table", then all
2575 entries are dumped. Since this can be quite heavy, it is possible to specify
2576 a filter in order to specify what entries to display.
2577
2578 When the "data." form is used the filter applies to the stored data (see
2579 "stick-table" in section 4.2). A stored data type must be specified
2580 in <type>, and this data type must be stored in the table otherwise an
2581 error is reported. The data is compared according to <operator> with the
2582 64-bit integer <value>. Operators are the same as with the ACLs :
2583
2584 - eq : match entries whose data is equal to this value
2585 - ne : match entries whose data is not equal to this value
2586 - le : match entries whose data is less than or equal to this value
2587 - ge : match entries whose data is greater than or equal to this value
2588 - lt : match entries whose data is less than this value
2589 - gt : match entries whose data is greater than this value
2590
Adis Nezirovic1a693fc2020-01-16 15:19:29 +01002591 In this form, you can use multiple data filter entries, up to a maximum
2592 defined during build time (4 by default).
Willy Tarreau44aed902015-10-13 14:45:29 +02002593
2594 When the key form is used the entry <key> is shown. The key must be of the
2595 same type as the table, which currently is limited to IPv4, IPv6, integer,
2596 and string.
2597
2598 Example :
2599 $ echo "show table http_proxy" | socat stdio /tmp/sock1
2600 >>> # table: http_proxy, type: ip, size:204800, used:2
2601 >>> 0x80e6a4c: key=127.0.0.1 use=0 exp=3594729 gpc0=0 conn_rate(30000)=1 \
2602 bytes_out_rate(60000)=187
2603 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
2604 bytes_out_rate(60000)=191
2605
2606 $ echo "show table http_proxy data.gpc0 gt 0" | socat stdio /tmp/sock1
2607 >>> # table: http_proxy, type: ip, size:204800, used:2
2608 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
2609 bytes_out_rate(60000)=191
2610
2611 $ echo "show table http_proxy data.conn_rate gt 5" | \
2612 socat stdio /tmp/sock1
2613 >>> # table: http_proxy, type: ip, size:204800, used:2
2614 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
2615 bytes_out_rate(60000)=191
2616
2617 $ echo "show table http_proxy key 127.0.0.2" | \
2618 socat stdio /tmp/sock1
2619 >>> # table: http_proxy, type: ip, size:204800, used:2
2620 >>> 0x80e6a80: key=127.0.0.2 use=0 exp=3594740 gpc0=1 conn_rate(30000)=10 \
2621 bytes_out_rate(60000)=191
2622
2623 When the data criterion applies to a dynamic value dependent on time such as
2624 a bytes rate, the value is dynamically computed during the evaluation of the
2625 entry in order to decide whether it has to be dumped or not. This means that
2626 such a filter could match for some time then not match anymore because as
2627 time goes, the average event rate drops.
2628
2629 It is possible to use this to extract lists of IP addresses abusing the
2630 service, in order to monitor them or even blacklist them in a firewall.
2631 Example :
2632 $ echo "show table http_proxy data.gpc0 gt 0" \
2633 | socat stdio /tmp/sock1 \
2634 | fgrep 'key=' | cut -d' ' -f2 | cut -d= -f2 > abusers-ip.txt
2635 ( or | awk '/key/{ print a[split($2,a,"=")]; }' )
2636
Willy Tarreau4e2b6462019-05-16 17:44:30 +02002637show threads
2638 Dumps some internal states and structures for each thread, that may be useful
2639 to help developers understand a problem. The output tries to be readable by
Willy Tarreauc7091d82019-05-17 10:08:49 +02002640 showing one block per thread. When haproxy is built with USE_THREAD_DUMP=1,
2641 an advanced dump mechanism involving thread signals is used so that each
2642 thread can dump its own state in turn. Without this option, the thread
2643 processing the command shows all its details but the other ones are less
Willy Tarreaue6a02fa2019-05-22 07:06:44 +02002644 detailed. A star ('*') is displayed in front of the thread handling the
2645 command. A right angle bracket ('>') may also be displayed in front of
2646 threads which didn't make any progress since last invocation of this command,
2647 indicating a bug in the code which must absolutely be reported. When this
2648 happens between two threads it usually indicates a deadlock. If a thread is
2649 alone, it's a different bug like a corrupted list. In all cases the process
2650 needs is not fully functional anymore and needs to be restarted.
2651
2652 The output format is purposely not documented so that it can easily evolve as
2653 new needs are identified, without having to maintain any form of backwards
2654 compatibility, and just like with "show activity", the values are meaningless
2655 without the code at hand.
Willy Tarreau4e2b6462019-05-16 17:44:30 +02002656
William Lallemandbb933462016-05-31 21:09:53 +02002657show tls-keys [id|*]
2658 Dump all loaded TLS ticket keys references. The TLS ticket key reference ID
2659 and the file from which the keys have been loaded is shown. Both of those
2660 can be used to update the TLS keys using "set ssl tls-key". If an ID is
2661 specified as parameter, it will dump the tickets, using * it will dump every
2662 keys from every references.
Willy Tarreau44aed902015-10-13 14:45:29 +02002663
Simon Horman6f6bb382017-01-04 09:37:26 +01002664show schema json
2665 Dump the schema used for the output of "show info json" and "show stat json".
2666
2667 The contains no extra whitespace in order to reduce the volume of output.
2668 For human consumption passing the output through a pretty printer may be
2669 helpful. Example :
2670
2671 $ echo "show schema json" | socat /var/run/haproxy.sock stdio | \
2672 python -m json.tool
2673
2674 The schema follows "JSON Schema" (json-schema.org) and accordingly
2675 verifiers may be used to verify the output of "show info json" and "show
2676 stat json" against the schema.
2677
Willy Tarreauf909c912019-08-22 20:06:04 +02002678show trace [<source>]
2679 Show the current trace status. For each source a line is displayed with a
2680 single-character status indicating if the trace is stopped, waiting, or
2681 running. The output sink used by the trace is indicated (or "none" if none
2682 was set), as well as the number of dropped events in this sink, followed by a
2683 brief description of the source. If a source name is specified, a detailed
2684 list of all events supported by the source, and their status for each action
2685 (report, start, pause, stop), indicated by a "+" if they are enabled, or a
2686 "-" otherwise. All these events are independent and an event might trigger
2687 a start without being reported and conversely.
Simon Horman6f6bb382017-01-04 09:37:26 +01002688
Willy Tarreau44aed902015-10-13 14:45:29 +02002689shutdown frontend <frontend>
2690 Completely delete the specified frontend. All the ports it was bound to will
2691 be released. It will not be possible to enable the frontend anymore after
2692 this operation. This is intended to be used in environments where stopping a
2693 proxy is not even imaginable but a misconfigured proxy must be fixed. That
2694 way it's possible to release the port and bind it into another process to
2695 restore operations. The frontend will not appear at all on the stats page
2696 once it is terminated.
2697
2698 The frontend may be specified either by its name or by its numeric ID,
2699 prefixed with a sharp ('#').
2700
2701 This command is restricted and can only be issued on sockets configured for
2702 level "admin".
2703
2704shutdown session <id>
2705 Immediately terminate the session matching the specified session identifier.
2706 This identifier is the first field at the beginning of the lines in the dumps
2707 of "show sess" (it corresponds to the session pointer). This can be used to
2708 terminate a long-running session without waiting for a timeout or when an
2709 endless transfer is ongoing. Such terminated sessions are reported with a 'K'
2710 flag in the logs.
2711
2712shutdown sessions server <backend>/<server>
2713 Immediately terminate all the sessions attached to the specified server. This
2714 can be used to terminate long-running sessions after a server is put into
2715 maintenance mode, for instance. Such terminated sessions are reported with a
2716 'K' flag in the logs.
2717
Willy Tarreauf909c912019-08-22 20:06:04 +02002718trace
2719 The "trace" command alone lists the trace sources, their current status, and
2720 their brief descriptions. It is only meant as a menu to enter next levels,
2721 see other "trace" commands below.
2722
2723trace 0
2724 Immediately stops all traces. This is made to be used as a quick solution
2725 to terminate a debugging session or as an emergency action to be used in case
2726 complex traces were enabled on multiple sources and impact the service.
2727
2728trace <source> event [ [+|-|!]<name> ]
2729 Without argument, this will list all the events supported by the designated
2730 source. They are prefixed with a "-" if they are not enabled, or a "+" if
2731 they are enabled. It is important to note that a single trace may be labelled
2732 with multiple events, and as long as any of the enabled events matches one of
2733 the events labelled on the trace, the event will be passed to the trace
2734 subsystem. For example, receiving an HTTP/2 frame of type HEADERS may trigger
2735 a frame event and a stream event since the frame creates a new stream. If
2736 either the frame event or the stream event are enabled for this source, the
2737 frame will be passed to the trace framework.
2738
2739 With an argument, it is possible to toggle the state of each event and
2740 individually enable or disable them. Two special keywords are supported,
2741 "none", which matches no event, and is used to disable all events at once,
2742 and "any" which matches all events, and is used to enable all events at
2743 once. Other events are specific to the event source. It is possible to
2744 enable one event by specifying its name, optionally prefixed with '+' for
2745 better readability. It is possible to disable one event by specifying its
2746 name prefixed by a '-' or a '!'.
2747
2748 One way to completely disable a trace source is to pass "event none", and
2749 this source will instantly be totally ignored.
2750
2751trace <source> level [<level>]
Willy Tarreau2ea549b2019-08-29 08:01:48 +02002752 Without argument, this will list all trace levels for this source, and the
Willy Tarreauf909c912019-08-22 20:06:04 +02002753 current one will be indicated by a star ('*') prepended in front of it. With
Willy Tarreau2ea549b2019-08-29 08:01:48 +02002754 an argument, this will change the trace level to the specified level. Detail
Willy Tarreauf909c912019-08-22 20:06:04 +02002755 levels are a form of filters that are applied before reporting the events.
Willy Tarreau2ea549b2019-08-29 08:01:48 +02002756 These filters are used to selectively include or exclude events depending on
2757 their level of importance. For example a developer might need to know
2758 precisely where in the code an HTTP header was considered invalid while the
2759 end user may not even care about this header's validity at all. There are
2760 currently 5 distinct levels for a trace :
Willy Tarreauf909c912019-08-22 20:06:04 +02002761
2762 user this will report information that are suitable for use by a
2763 regular haproxy user who wants to observe his traffic.
2764 Typically some HTTP requests and responses will be reported
2765 without much detail. Most sources will set this as the
2766 default level to ease operations.
2767
Willy Tarreau2ea549b2019-08-29 08:01:48 +02002768 proto in addition to what is reported at the "user" level, it also
2769 displays protocol-level updates. This can for example be the
2770 frame types or HTTP headers after decoding.
Willy Tarreauf909c912019-08-22 20:06:04 +02002771
2772 state in addition to what is reported at the "proto" level, it
2773 will also display state transitions (or failed transitions)
2774 which happen in parsers, so this will show attempts to
2775 perform an operation while the "proto" level only shows
2776 the final operation.
2777
Willy Tarreau2ea549b2019-08-29 08:01:48 +02002778 data in addition to what is reported at the "state" level, it
2779 will also include data transfers between the various layers.
2780
Willy Tarreauf909c912019-08-22 20:06:04 +02002781 developer it reports everything available, which can include advanced
2782 information such as "breaking out of this loop" that are
2783 only relevant to a developer trying to understand a bug that
Willy Tarreau09fb0df2019-08-29 08:40:59 +02002784 only happens once in a while in field. Function names are
2785 only reported at this level.
Willy Tarreauf909c912019-08-22 20:06:04 +02002786
2787 It is highly recommended to always use the "user" level only and switch to
2788 other levels only if instructed to do so by a developer. Also it is a good
2789 idea to first configure the events before switching to higher levels, as it
2790 may save from dumping many lines if no filter is applied.
2791
2792trace <source> lock [criterion]
2793 Without argument, this will list all the criteria supported by this source
2794 for lock-on processing, and display the current choice by a star ('*') in
2795 front of it. Lock-on means that the source will focus on the first matching
2796 event and only stick to the criterion which triggered this event, and ignore
2797 all other ones until the trace stops. This allows for example to take a trace
2798 on a single connection or on a single stream. The following criteria are
2799 supported by some traces, though not necessarily all, since some of them
2800 might not be available to the source :
2801
2802 backend lock on the backend that started the trace
2803 connection lock on the connection that started the trace
2804 frontend lock on the frontend that started the trace
2805 listener lock on the listener that started the trace
2806 nothing do not lock on anything
2807 server lock on the server that started the trace
2808 session lock on the session that started the trace
2809 thread lock on the thread that started the trace
2810
2811 In addition to this, each source may provide up to 4 specific criteria such
2812 as internal states or connection IDs. For example in HTTP/2 it is possible
2813 to lock on the H2 stream and ignore other streams once a strace starts.
2814
2815 When a criterion is passed in argument, this one is used instead of the
2816 other ones and any existing tracking is immediately terminated so that it can
2817 restart with the new criterion. The special keyword "nothing" is supported by
2818 all sources to permanently disable tracking.
2819
2820trace <source> { pause | start | stop } [ [+|-|!]event]
2821 Without argument, this will list the events enabled to automatically pause,
2822 start, or stop a trace for this source. These events are specific to each
2823 trace source. With an argument, this will either enable the event for the
2824 specified action (if optionally prefixed by a '+') or disable it (if
2825 prefixed by a '-' or '!'). The special keyword "now" is not an event and
2826 requests to take the action immediately. The keywords "none" and "any" are
2827 supported just like in "trace event".
2828
2829 The 3 supported actions are respectively "pause", "start" and "stop". The
2830 "pause" action enumerates events which will cause a running trace to stop and
2831 wait for a new start event to restart it. The "start" action enumerates the
2832 events which switch the trace into the waiting mode until one of the start
2833 events appears. And the "stop" action enumerates the events which definitely
2834 stop the trace until it is manually enabled again. In practice it makes sense
2835 to manually start a trace using "start now" without caring about events, and
2836 to stop it using "stop now". In order to capture more subtle event sequences,
2837 setting "start" to a normal event (like receiving an HTTP request) and "stop"
2838 to a very rare event like emitting a certain error, will ensure that the last
2839 captured events will match the desired criteria. And the pause event is
2840 useful to detect the end of a sequence, disable the lock-on and wait for
2841 another opportunity to take a capture. In this case it can make sense to
2842 enable lock-on to spot only one specific criterion (e.g. a stream), and have
2843 "start" set to anything that starts this criterion (e.g. all events which
2844 create a stream), "stop" set to the expected anomaly, and "pause" to anything
2845 that ends that criterion (e.g. any end of stream event). In this case the
2846 trace log will contain complete sequences of perfectly clean series affecting
2847 a single object, until the last sequence containing everything from the
2848 beginning to the anomaly.
2849
2850trace <source> sink [<sink>]
2851 Without argument, this will list all event sinks available for this source,
2852 and the currently configured one will have a star ('*') prepended in front
2853 of it. Sink "none" is always available and means that all events are simply
2854 dropped, though their processing is not ignored (e.g. lock-on does occur).
2855 Other sinks are available depending on configuration and build options, but
2856 typically "stdout" and "stderr" will be usable in debug mode, and in-memory
2857 ring buffers should be available as well. When a name is specified, the sink
2858 instantly changes for the specified source. Events are not changed during a
2859 sink change. In the worst case some may be lost if an invalid sink is used
2860 (or "none"), but operations do continue to a different destination.
2861
Willy Tarreau370a6942019-08-29 08:24:16 +02002862trace <source> verbosity [<level>]
2863 Without argument, this will list all verbosity levels for this source, and the
2864 current one will be indicated by a star ('*') prepended in front of it. With
2865 an argument, this will change the verbosity level to the specified one.
2866
2867 Verbosity levels indicate how far the trace decoder should go to provide
2868 detailed information. It depends on the trace source, since some sources will
2869 not even provide a specific decoder. Level "quiet" is always available and
2870 disables any decoding. It can be useful when trying to figure what's
2871 happening before trying to understand the details, since it will have a very
2872 low impact on performance and trace size. When no verbosity levels are
2873 declared by a source, level "default" is available and will cause a decoder
2874 to be called when specified in the traces. It is an opportunistic decoding.
2875 When the source declares some verbosity levels, these ones are listed with
2876 a description of what they correspond to. In this case the trace decoder
2877 provided by the source will be as accurate as possible based on the
2878 information available at the trace point. The first level above "quiet" is
2879 set by default.
2880
Willy Tarreau2212e6a2015-10-13 14:40:55 +02002881
William Lallemand142db372018-12-11 18:56:45 +010028829.4. Master CLI
2883---------------
2884
2885The master CLI is a socket bound to the master process in master-worker mode.
2886This CLI gives access to the unix socket commands in every running or leaving
2887processes and allows a basic supervision of those processes.
2888
2889The master CLI is configurable only from the haproxy program arguments with
2890the -S option. This option also takes bind options separated by commas.
2891
2892Example:
2893
2894 # haproxy -W -S 127.0.0.1:1234 -f test1.cfg
2895 # haproxy -Ws -S /tmp/master-socket,uid,1000,gid,1000,mode,600 -f test1.cfg
William Lallemandb7ea1412018-12-13 09:05:47 +01002896 # haproxy -W -S /tmp/master-socket,level,user -f test1.cfg
William Lallemand142db372018-12-11 18:56:45 +01002897
2898The master CLI introduces a new 'show proc' command to surpervise the
2899processes:
2900
2901Example:
2902
2903 $ echo 'show proc' | socat /var/run/haproxy-master.sock -
William Lallemand1dc69632019-06-12 19:11:33 +02002904 #<PID> <type> <relative PID> <reloads> <uptime> <version>
2905 1162 master 0 5 0d00h02m07s 2.0-dev7-0124c9-7
William Lallemand142db372018-12-11 18:56:45 +01002906 # workers
William Lallemand1dc69632019-06-12 19:11:33 +02002907 1271 worker 1 0 0d00h00m00s 2.0-dev7-0124c9-7
2908 1272 worker 2 0 0d00h00m00s 2.0-dev7-0124c9-7
William Lallemand142db372018-12-11 18:56:45 +01002909 # old workers
William Lallemand1dc69632019-06-12 19:11:33 +02002910 1233 worker [was: 1] 3 0d00h00m43s 2.0-dev3-6019f6-289
William Lallemand142db372018-12-11 18:56:45 +01002911
2912
2913In this example, the master has been reloaded 5 times but one of the old
2914worker is still running and survived 3 reloads. You could access the CLI of
2915this worker to understand what's going on.
2916
Willy Tarreau52880f92018-12-15 13:30:03 +01002917When the prompt is enabled (via the "prompt" command), the context the CLI is
2918working on is displayed in the prompt. The master is identified by the "master"
2919string, and other processes are identified with their PID. In case the last
2920reload failed, the master prompt will be changed to "master[ReloadFailed]>" so
2921that it becomes visible that the process is still running on the previous
2922configuration and that the new configuration is not operational.
2923
William Lallemand142db372018-12-11 18:56:45 +01002924The master CLI uses a special prefix notation to access the multiple
2925processes. This notation is easily identifiable as it begins by a @.
2926
2927A @ prefix can be followed by a relative process number or by an exclamation
2928point and a PID. (e.g. @1 or @!1271). A @ alone could be use to specify the
2929master. Leaving processes are only accessible with the PID as relative process
2930number are only usable with the current processes.
2931
2932Examples:
2933
2934 $ socat /var/run/haproxy-master.sock readline
2935 prompt
2936 master> @1 show info; @2 show info
2937 [...]
2938 Process_num: 1
2939 Pid: 1271
2940 [...]
2941 Process_num: 2
2942 Pid: 1272
2943 [...]
2944 master>
2945
2946 $ echo '@!1271 show info; @!1272 show info' | socat /var/run/haproxy-master.sock -
2947 [...]
2948
2949A prefix could be use as a command, which will send every next commands to
2950the specified process.
2951
2952Examples:
2953
2954 $ socat /var/run/haproxy-master.sock readline
2955 prompt
2956 master> @1
2957 1271> show info
2958 [...]
2959 1271> show stat
2960 [...]
2961 1271> @
2962 master>
2963
2964 $ echo '@1; show info; show stat; @2; show info; show stat' | socat /var/run/haproxy-master.sock -
2965 [...]
2966
William Lallemanda57b7e32018-12-14 21:11:31 +01002967You can also reload the HAProxy master process with the "reload" command which
2968does the same as a `kill -USR2` on the master process, provided that the user
2969has at least "operator" or "admin" privileges.
2970
2971Example:
2972
2973 $ echo "reload" | socat /var/run/haproxy-master.sock
2974
2975Note that a reload will close the connection to the master CLI.
2976
William Lallemand142db372018-12-11 18:56:45 +01002977
Willy Tarreau2212e6a2015-10-13 14:40:55 +0200297810. Tricks for easier configuration management
2979----------------------------------------------
2980
2981It is very common that two HAProxy nodes constituting a cluster share exactly
2982the same configuration modulo a few addresses. Instead of having to maintain a
2983duplicate configuration for each node, which will inevitably diverge, it is
2984possible to include environment variables in the configuration. Thus multiple
2985configuration may share the exact same file with only a few different system
2986wide environment variables. This started in version 1.5 where only addresses
2987were allowed to include environment variables, and 1.6 goes further by
2988supporting environment variables everywhere. The syntax is the same as in the
2989UNIX shell, a variable starts with a dollar sign ('$'), followed by an opening
2990curly brace ('{'), then the variable name followed by the closing brace ('}').
2991Except for addresses, environment variables are only interpreted in arguments
2992surrounded with double quotes (this was necessary not to break existing setups
2993using regular expressions involving the dollar symbol).
2994
2995Environment variables also make it convenient to write configurations which are
2996expected to work on various sites where only the address changes. It can also
2997permit to remove passwords from some configs. Example below where the the file
2998"site1.env" file is sourced by the init script upon startup :
2999
3000 $ cat site1.env
3001 LISTEN=192.168.1.1
3002 CACHE_PFX=192.168.11
3003 SERVER_PFX=192.168.22
3004 LOGGER=192.168.33.1
3005 STATSLP=admin:pa$$w0rd
3006 ABUSERS=/etc/haproxy/abuse.lst
3007 TIMEOUT=10s
3008
3009 $ cat haproxy.cfg
3010 global
3011 log "${LOGGER}:514" local0
3012
3013 defaults
3014 mode http
3015 timeout client "${TIMEOUT}"
3016 timeout server "${TIMEOUT}"
3017 timeout connect 5s
3018
3019 frontend public
3020 bind "${LISTEN}:80"
3021 http-request reject if { src -f "${ABUSERS}" }
3022 stats uri /stats
3023 stats auth "${STATSLP}"
3024 use_backend cache if { path_end .jpg .css .ico }
3025 default_backend server
3026
3027 backend cache
3028 server cache1 "${CACHE_PFX}.1:18080" check
3029 server cache2 "${CACHE_PFX}.2:18080" check
3030
3031 backend server
3032 server cache1 "${SERVER_PFX}.1:8080" check
3033 server cache2 "${SERVER_PFX}.2:8080" check
3034
3035
303611. Well-known traps to avoid
3037-----------------------------
3038
3039Once in a while, someone reports that after a system reboot, the haproxy
3040service wasn't started, and that once they start it by hand it works. Most
3041often, these people are running a clustered IP address mechanism such as
3042keepalived, to assign the service IP address to the master node only, and while
3043it used to work when they used to bind haproxy to address 0.0.0.0, it stopped
3044working after they bound it to the virtual IP address. What happens here is
3045that when the service starts, the virtual IP address is not yet owned by the
3046local node, so when HAProxy wants to bind to it, the system rejects this
3047because it is not a local IP address. The fix doesn't consist in delaying the
3048haproxy service startup (since it wouldn't stand a restart), but instead to
3049properly configure the system to allow binding to non-local addresses. This is
3050easily done on Linux by setting the net.ipv4.ip_nonlocal_bind sysctl to 1. This
3051is also needed in order to transparently intercept the IP traffic that passes
3052through HAProxy for a specific target address.
3053
3054Multi-process configurations involving source port ranges may apparently seem
3055to work but they will cause some random failures under high loads because more
3056than one process may try to use the same source port to connect to the same
3057server, which is not possible. The system will report an error and a retry will
3058happen, picking another port. A high value in the "retries" parameter may hide
3059the effect to a certain extent but this also comes with increased CPU usage and
3060processing time. Logs will also report a certain number of retries. For this
3061reason, port ranges should be avoided in multi-process configurations.
3062
Dan Lloyd8e48b872016-07-01 21:01:18 -04003063Since HAProxy uses SO_REUSEPORT and supports having multiple independent
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003064processes bound to the same IP:port, during troubleshooting it can happen that
3065an old process was not stopped before a new one was started. This provides
3066absurd test results which tend to indicate that any change to the configuration
3067is ignored. The reason is that in fact even the new process is restarted with a
3068new configuration, the old one also gets some incoming connections and
3069processes them, returning unexpected results. When in doubt, just stop the new
3070process and try again. If it still works, it very likely means that an old
3071process remains alive and has to be stopped. Linux's "netstat -lntp" is of good
3072help here.
3073
3074When adding entries to an ACL from the command line (eg: when blacklisting a
3075source address), it is important to keep in mind that these entries are not
3076synchronized to the file and that if someone reloads the configuration, these
3077updates will be lost. While this is often the desired effect (for blacklisting)
3078it may not necessarily match expectations when the change was made as a fix for
3079a problem. See the "add acl" action of the CLI interface.
3080
3081
308212. Debugging and performance issues
3083------------------------------------
3084
3085When HAProxy is started with the "-d" option, it will stay in the foreground
3086and will print one line per event, such as an incoming connection, the end of a
3087connection, and for each request or response header line seen. This debug
3088output is emitted before the contents are processed, so they don't consider the
3089local modifications. The main use is to show the request and response without
3090having to run a network sniffer. The output is less readable when multiple
3091connections are handled in parallel, though the "debug2ansi" and "debug2html"
3092scripts found in the examples/ directory definitely help here by coloring the
3093output.
3094
3095If a request or response is rejected because HAProxy finds it is malformed, the
3096best thing to do is to connect to the CLI and issue "show errors", which will
3097report the last captured faulty request and response for each frontend and
3098backend, with all the necessary information to indicate precisely the first
3099character of the input stream that was rejected. This is sometimes needed to
3100prove to customers or to developers that a bug is present in their code. In
3101this case it is often possible to relax the checks (but still keep the
3102captures) using "option accept-invalid-http-request" or its equivalent for
3103responses coming from the server "option accept-invalid-http-response". Please
3104see the configuration manual for more details.
3105
3106Example :
3107
3108 > show errors
3109 Total events captured on [13/Oct/2015:13:43:47.169] : 1
3110
3111 [13/Oct/2015:13:43:40.918] frontend HAProxyLocalStats (#2): invalid request
3112 backend <NONE> (#-1), server <NONE> (#-1), event #0
3113 src 127.0.0.1:51981, session #0, session flags 0x00000080
3114 HTTP msg state 26, msg flags 0x00000000, tx flags 0x00000000
3115 HTTP chunk len 0 bytes, HTTP body len 0 bytes
3116 buffer flags 0x00808002, out 0 bytes, total 31 bytes
3117 pending 31 bytes, wrapping at 8040, error at position 13:
3118
3119 00000 GET /invalid request HTTP/1.1\r\n
3120
3121
3122The output of "show info" on the CLI provides a number of useful information
3123regarding the maximum connection rate ever reached, maximum SSL key rate ever
3124reached, and in general all information which can help to explain temporary
3125issues regarding CPU or memory usage. Example :
3126
3127 > show info
3128 Name: HAProxy
3129 Version: 1.6-dev7-e32d18-17
3130 Release_date: 2015/10/12
3131 Nbproc: 1
3132 Process_num: 1
3133 Pid: 7949
3134 Uptime: 0d 0h02m39s
3135 Uptime_sec: 159
3136 Memmax_MB: 0
3137 Ulimit-n: 120032
3138 Maxsock: 120032
3139 Maxconn: 60000
3140 Hard_maxconn: 60000
3141 CurrConns: 0
3142 CumConns: 3
3143 CumReq: 3
3144 MaxSslConns: 0
3145 CurrSslConns: 0
3146 CumSslConns: 0
3147 Maxpipes: 0
3148 PipesUsed: 0
3149 PipesFree: 0
3150 ConnRate: 0
3151 ConnRateLimit: 0
3152 MaxConnRate: 1
3153 SessRate: 0
3154 SessRateLimit: 0
3155 MaxSessRate: 1
3156 SslRate: 0
3157 SslRateLimit: 0
3158 MaxSslRate: 0
3159 SslFrontendKeyRate: 0
3160 SslFrontendMaxKeyRate: 0
3161 SslFrontendSessionReuse_pct: 0
3162 SslBackendKeyRate: 0
3163 SslBackendMaxKeyRate: 0
3164 SslCacheLookups: 0
3165 SslCacheMisses: 0
3166 CompressBpsIn: 0
3167 CompressBpsOut: 0
3168 CompressBpsRateLim: 0
3169 ZlibMemUsage: 0
3170 MaxZlibMemUsage: 0
3171 Tasks: 5
3172 Run_queue: 1
3173 Idle_pct: 100
3174 node: wtap
3175 description:
3176
3177When an issue seems to randomly appear on a new version of HAProxy (eg: every
3178second request is aborted, occasional crash, etc), it is worth trying to enable
Dan Lloyd8e48b872016-07-01 21:01:18 -04003179memory poisoning so that each call to malloc() is immediately followed by the
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003180filling of the memory area with a configurable byte. By default this byte is
31810x50 (ASCII for 'P'), but any other byte can be used, including zero (which
3182will have the same effect as a calloc() and which may make issues disappear).
Dan Lloyd8e48b872016-07-01 21:01:18 -04003183Memory poisoning is enabled on the command line using the "-dM" option. It
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003184slightly hurts performance and is not recommended for use in production. If
Dan Lloyd8e48b872016-07-01 21:01:18 -04003185an issue happens all the time with it or never happens when poisoning uses
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003186byte zero, it clearly means you've found a bug and you definitely need to
3187report it. Otherwise if there's no clear change, the problem it is not related.
3188
3189When debugging some latency issues, it is important to use both strace and
3190tcpdump on the local machine, and another tcpdump on the remote system. The
3191reason for this is that there are delays everywhere in the processing chain and
3192it is important to know which one is causing latency to know where to act. In
3193practice, the local tcpdump will indicate when the input data come in. Strace
3194will indicate when haproxy receives these data (using recv/recvfrom). Warning,
3195openssl uses read()/write() syscalls instead of recv()/send(). Strace will also
3196show when haproxy sends the data, and tcpdump will show when the system sends
3197these data to the interface. Then the external tcpdump will show when the data
3198sent are really received (since the local one only shows when the packets are
3199queued). The benefit of sniffing on the local system is that strace and tcpdump
3200will use the same reference clock. Strace should be used with "-tts200" to get
3201complete timestamps and report large enough chunks of data to read them.
3202Tcpdump should be used with "-nvvttSs0" to report full packets, real sequence
3203numbers and complete timestamps.
3204
3205In practice, received data are almost always immediately received by haproxy
3206(unless the machine has a saturated CPU or these data are invalid and not
3207delivered). If these data are received but not sent, it generally is because
3208the output buffer is saturated (ie: recipient doesn't consume the data fast
3209enough). This can be confirmed by seeing that the polling doesn't notify of
3210the ability to write on the output file descriptor for some time (it's often
3211easier to spot in the strace output when the data finally leave and then roll
3212back to see when the write event was notified). It generally matches an ACK
3213received from the recipient, and detected by tcpdump. Once the data are sent,
3214they may spend some time in the system doing nothing. Here again, the TCP
3215congestion window may be limited and not allow these data to leave, waiting for
3216an ACK to open the window. If the traffic is idle and the data take 40 ms or
3217200 ms to leave, it's a different issue (which is not an issue), it's the fact
3218that the Nagle algorithm prevents empty packets from leaving immediately, in
3219hope that they will be merged with subsequent data. HAProxy automatically
3220disables Nagle in pure TCP mode and in tunnels. However it definitely remains
3221enabled when forwarding an HTTP body (and this contributes to the performance
3222improvement there by reducing the number of packets). Some HTTP non-compliant
3223applications may be sensitive to the latency when delivering incomplete HTTP
3224response messages. In this case you will have to enable "option http-no-delay"
3225to disable Nagle in order to work around their design, keeping in mind that any
3226other proxy in the chain may similarly be impacted. If tcpdump reports that data
3227leave immediately but the other end doesn't see them quickly, it can mean there
Dan Lloyd8e48b872016-07-01 21:01:18 -04003228is a congested WAN link, a congested LAN with flow control enabled and
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003229preventing the data from leaving, or more commonly that HAProxy is in fact
3230running in a virtual machine and that for whatever reason the hypervisor has
3231decided that the data didn't need to be sent immediately. In virtualized
3232environments, latency issues are almost always caused by the virtualization
3233layer, so in order to save time, it's worth first comparing tcpdump in the VM
3234and on the external components. Any difference has to be credited to the
3235hypervisor and its accompanying drivers.
3236
3237When some TCP SACK segments are seen in tcpdump traces (using -vv), it always
3238means that the side sending them has got the proof of a lost packet. While not
3239seeing them doesn't mean there are no losses, seeing them definitely means the
3240network is lossy. Losses are normal on a network, but at a rate where SACKs are
3241not noticeable at the naked eye. If they appear a lot in the traces, it is
3242worth investigating exactly what happens and where the packets are lost. HTTP
3243doesn't cope well with TCP losses, which introduce huge latencies.
3244
3245The "netstat -i" command will report statistics per interface. An interface
3246where the Rx-Ovr counter grows indicates that the system doesn't have enough
3247resources to receive all incoming packets and that they're lost before being
3248processed by the network driver. Rx-Drp indicates that some received packets
3249were lost in the network stack because the application doesn't process them
3250fast enough. This can happen during some attacks as well. Tx-Drp means that
3251the output queues were full and packets had to be dropped. When using TCP it
Dan Lloyd8e48b872016-07-01 21:01:18 -04003252should be very rare, but will possibly indicate a saturated outgoing link.
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003253
3254
325513. Security considerations
3256---------------------------
3257
3258HAProxy is designed to run with very limited privileges. The standard way to
3259use it is to isolate it into a chroot jail and to drop its privileges to a
3260non-root user without any permissions inside this jail so that if any future
3261vulnerability were to be discovered, its compromise would not affect the rest
3262of the system.
3263
Dan Lloyd8e48b872016-07-01 21:01:18 -04003264In order to perform a chroot, it first needs to be started as a root user. It is
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003265pointless to build hand-made chroots to start the process there, these ones are
3266painful to build, are never properly maintained and always contain way more
3267bugs than the main file-system. And in case of compromise, the intruder can use
3268the purposely built file-system. Unfortunately many administrators confuse
3269"start as root" and "run as root", resulting in the uid change to be done prior
3270to starting haproxy, and reducing the effective security restrictions.
3271
3272HAProxy will need to be started as root in order to :
3273 - adjust the file descriptor limits
3274 - bind to privileged port numbers
3275 - bind to a specific network interface
3276 - transparently listen to a foreign address
3277 - isolate itself inside the chroot jail
3278 - drop to another non-privileged UID
3279
3280HAProxy may require to be run as root in order to :
3281 - bind to an interface for outgoing connections
3282 - bind to privileged source ports for outgoing connections
Dan Lloyd8e48b872016-07-01 21:01:18 -04003283 - transparently bind to a foreign address for outgoing connections
Willy Tarreau2212e6a2015-10-13 14:40:55 +02003284
3285Most users will never need the "run as root" case. But the "start as root"
3286covers most usages.
3287
3288A safe configuration will have :
3289
3290 - a chroot statement pointing to an empty location without any access
3291 permissions. This can be prepared this way on the UNIX command line :
3292
3293 # mkdir /var/empty && chmod 0 /var/empty || echo "Failed"
3294
3295 and referenced like this in the HAProxy configuration's global section :
3296
3297 chroot /var/empty
3298
3299 - both a uid/user and gid/group statements in the global section :
3300
3301 user haproxy
3302 group haproxy
3303
3304 - a stats socket whose mode, uid and gid are set to match the user and/or
3305 group allowed to access the CLI so that nobody may access it :
3306
3307 stats socket /var/run/haproxy.stat uid hatop gid hatop mode 600
3308