blob: b521bf35aabbdd9360daa0804a9eb712f9923c6c [file] [log] [blame]
----------------------
HAProxy
Configuration Manual
----------------------
version 1.5
willy tarreau
2013/06/17
This document covers the configuration language as implemented in the version
specified above. It does not provide any hint, example or advice. For such
documentation, please refer to the Reference Manual or the Architecture Manual.
The summary below is meant to help you search sections by name and navigate
through the document.
Note to documentation contributors :
This document is formatted with 80 columns per line, with even number of
spaces for indentation and without tabs. Please follow these rules strictly
so that it remains easily printable everywhere. If a line needs to be
printed verbatim and does not fit, please end each line with a backslash
('\') and continue on next line, indented by two characters. It is also
sometimes useful to prefix all output lines (logs, console outs) with 3
closing angle brackets ('>>>') in order to help get the difference between
inputs and outputs when it can become ambiguous. If you add sections,
please update the summary below for easier searching.
Summary
-------
1. Quick reminder about HTTP
1.1. The HTTP transaction model
1.2. HTTP request
1.2.1. The Request line
1.2.2. The request headers
1.3. HTTP response
1.3.1. The Response line
1.3.2. The response headers
2. Configuring HAProxy
2.1. Configuration file format
2.2. Time format
2.3. Examples
3. Global parameters
3.1. Process management and security
3.2. Performance tuning
3.3. Debugging
3.4. Userlists
3.5. Peers
4. Proxies
4.1. Proxy keywords matrix
4.2. Alphabetically sorted keywords reference
5. Bind and Server options
5.1. Bind options
5.2. Server and default-server options
6. HTTP header manipulation
7. Using ACLs and fetching samples
7.1. ACL basics
7.1.1. Matching booleans
7.1.2. Matching integers
7.1.3. Matching strings
7.1.4. Matching regular expressions (regexes)
7.1.5. Matching arbitrary data blocks
7.1.6. Matching IPv4 and IPv6 addresses
7.2. Using ACLs to form conditions
7.3. Fetching samples
7.3.1. Fetching samples from internal states
7.3.2. Fetching samples at Layer 4
7.3.3. Fetching samples at Layer 5
7.3.4. Fetching samples from buffer contents (Layer 6)
7.3.5. Fetching HTTP samples (Layer 7)
7.4. Pre-defined ACLs
8. Logging
8.1. Log levels
8.2. Log formats
8.2.1. Default log format
8.2.2. TCP log format
8.2.3. HTTP log format
8.2.4. Custom log format
8.2.5. Error log format
8.3. Advanced logging options
8.3.1. Disabling logging of external tests
8.3.2. Logging before waiting for the session to terminate
8.3.3. Raising log level upon errors
8.3.4. Disabling logging of successful connections
8.4. Timing events
8.5. Session state at disconnection
8.6. Non-printable characters
8.7. Capturing HTTP cookies
8.8. Capturing HTTP headers
8.9. Examples of logs
9. Statistics and monitoring
9.1. CSV format
9.2. Unix Socket commands
1. Quick reminder about HTTP
----------------------------
When haproxy is running in HTTP mode, both the request and the response are
fully analyzed and indexed, thus it becomes possible to build matching criteria
on almost anything found in the contents.
However, it is important to understand how HTTP requests and responses are
formed, and how HAProxy decomposes them. It will then become easier to write
correct rules and to debug existing configurations.
1.1. The HTTP transaction model
-------------------------------
The HTTP protocol is transaction-driven. This means that each request will lead
to one and only one response. Traditionally, a TCP connection is established
from the client to the server, a request is sent by the client on the
connection, the server responds and the connection is closed. A new request
will involve a new connection :
[CON1] [REQ1] ... [RESP1] [CLO1] [CON2] [REQ2] ... [RESP2] [CLO2] ...
In this mode, called the "HTTP close" mode, there are as many connection
establishments as there are HTTP transactions. Since the connection is closed
by the server after the response, the client does not need to know the content
length.
Due to the transactional nature of the protocol, it was possible to improve it
to avoid closing a connection between two subsequent transactions. In this mode
however, it is mandatory that the server indicates the content length for each
response so that the client does not wait indefinitely. For this, a special
header is used: "Content-length". This mode is called the "keep-alive" mode :
[CON] [REQ1] ... [RESP1] [REQ2] ... [RESP2] [CLO] ...
Its advantages are a reduced latency between transactions, and less processing
power required on the server side. It is generally better than the close mode,
but not always because the clients often limit their concurrent connections to
a smaller value.
A last improvement in the communications is the pipelining mode. It still uses
keep-alive, but the client does not wait for the first response to send the
second request. This is useful for fetching large number of images composing a
page :
[CON] [REQ1] [REQ2] ... [RESP1] [RESP2] [CLO] ...
This can obviously have a tremendous benefit on performance because the network
latency is eliminated between subsequent requests. Many HTTP agents do not
correctly support pipelining since there is no way to associate a response with
the corresponding request in HTTP. For this reason, it is mandatory for the
server to reply in the exact same order as the requests were received.
By default HAProxy operates in a tunnel-like mode with regards to persistent
connections: for each connection it processes the first request and forwards
everything else (including additional requests) to selected server. Once
established, the connection is persisted both on the client and server
sides. Use "option http-server-close" to preserve client persistent connections
while handling every incoming request individually, dispatching them one after
another to servers, in HTTP close mode. Use "option httpclose" to switch both
sides to HTTP close mode. "option forceclose" and "option
http-pretend-keepalive" help working around servers misbehaving in HTTP close
mode.
1.2. HTTP request
-----------------
First, let's consider this HTTP request :
Line Contents
number
1 GET /serv/login.php?lang=en&profile=2 HTTP/1.1
2 Host: www.mydomain.com
3 User-agent: my small browser
4 Accept: image/jpeg, image/gif
5 Accept: image/png
1.2.1. The Request line
-----------------------
Line 1 is the "request line". It is always composed of 3 fields :
- a METHOD : GET
- a URI : /serv/login.php?lang=en&profile=2
- a version tag : HTTP/1.1
All of them are delimited by what the standard calls LWS (linear white spaces),
which are commonly spaces, but can also be tabs or line feeds/carriage returns
followed by spaces/tabs. The method itself cannot contain any colon (':') and
is limited to alphabetic letters. All those various combinations make it
desirable that HAProxy performs the splitting itself rather than leaving it to
the user to write a complex or inaccurate regular expression.
The URI itself can have several forms :
- A "relative URI" :
/serv/login.php?lang=en&profile=2
It is a complete URL without the host part. This is generally what is
received by servers, reverse proxies and transparent proxies.
- An "absolute URI", also called a "URL" :
http://192.168.0.12:8080/serv/login.php?lang=en&profile=2
It is composed of a "scheme" (the protocol name followed by '://'), a host
name or address, optionally a colon (':') followed by a port number, then
a relative URI beginning at the first slash ('/') after the address part.
This is generally what proxies receive, but a server supporting HTTP/1.1
must accept this form too.
- a star ('*') : this form is only accepted in association with the OPTIONS
method and is not relayable. It is used to inquiry a next hop's
capabilities.
- an address:port combination : 192.168.0.12:80
This is used with the CONNECT method, which is used to establish TCP
tunnels through HTTP proxies, generally for HTTPS, but sometimes for
other protocols too.
In a relative URI, two sub-parts are identified. The part before the question
mark is called the "path". It is typically the relative path to static objects
on the server. The part after the question mark is called the "query string".
It is mostly used with GET requests sent to dynamic scripts and is very
specific to the language, framework or application in use.
1.2.2. The request headers
--------------------------
The headers start at the second line. They are composed of a name at the
beginning of the line, immediately followed by a colon (':'). Traditionally,
an LWS is added after the colon but that's not required. Then come the values.
Multiple identical headers may be folded into one single line, delimiting the
values with commas, provided that their order is respected. This is commonly
encountered in the "Cookie:" field. A header may span over multiple lines if
the subsequent lines begin with an LWS. In the example in 1.2, lines 4 and 5
define a total of 3 values for the "Accept:" header.
Contrary to a common mis-conception, header names are not case-sensitive, and
their values are not either if they refer to other header names (such as the
"Connection:" header).
The end of the headers is indicated by the first empty line. People often say
that it's a double line feed, which is not exact, even if a double line feed
is one valid form of empty line.
Fortunately, HAProxy takes care of all these complex combinations when indexing
headers, checking values and counting them, so there is no reason to worry
about the way they could be written, but it is important not to accuse an
application of being buggy if it does unusual, valid things.
Important note:
As suggested by RFC2616, HAProxy normalizes headers by replacing line breaks
in the middle of headers by LWS in order to join multi-line headers. This
is necessary for proper analysis and helps less capable HTTP parsers to work
correctly and not to be fooled by such complex constructs.
1.3. HTTP response
------------------
An HTTP response looks very much like an HTTP request. Both are called HTTP
messages. Let's consider this HTTP response :
Line Contents
number
1 HTTP/1.1 200 OK
2 Content-length: 350
3 Content-Type: text/html
As a special case, HTTP supports so called "Informational responses" as status
codes 1xx. These messages are special in that they don't convey any part of the
response, they're just used as sort of a signaling message to ask a client to
continue to post its request for instance. In the case of a status 100 response
the requested information will be carried by the next non-100 response message
following the informational one. This implies that multiple responses may be
sent to a single request, and that this only works when keep-alive is enabled
(1xx messages are HTTP/1.1 only). HAProxy handles these messages and is able to
correctly forward and skip them, and only process the next non-100 response. As
such, these messages are neither logged nor transformed, unless explicitly
state otherwise. Status 101 messages indicate that the protocol is changing
over the same connection and that haproxy must switch to tunnel mode, just as
if a CONNECT had occurred. Then the Upgrade header would contain additional
information about the type of protocol the connection is switching to.
1.3.1. The Response line
------------------------
Line 1 is the "response line". It is always composed of 3 fields :
- a version tag : HTTP/1.1
- a status code : 200
- a reason : OK
The status code is always 3-digit. The first digit indicates a general status :
- 1xx = informational message to be skipped (eg: 100, 101)
- 2xx = OK, content is following (eg: 200, 206)
- 3xx = OK, no content following (eg: 302, 304)
- 4xx = error caused by the client (eg: 401, 403, 404)
- 5xx = error caused by the server (eg: 500, 502, 503)
Please refer to RFC2616 for the detailed meaning of all such codes. The
"reason" field is just a hint, but is not parsed by clients. Anything can be
found there, but it's a common practice to respect the well-established
messages. It can be composed of one or multiple words, such as "OK", "Found",
or "Authentication Required".
Haproxy may emit the following status codes by itself :
Code When / reason
200 access to stats page, and when replying to monitoring requests
301 when performing a redirection, depending on the configured code
302 when performing a redirection, depending on the configured code
303 when performing a redirection, depending on the configured code
307 when performing a redirection, depending on the configured code
308 when performing a redirection, depending on the configured code
400 for an invalid or too large request
401 when an authentication is required to perform the action (when
accessing the stats page)
403 when a request is forbidden by a "block" ACL or "reqdeny" filter
408 when the request timeout strikes before the request is complete
500 when haproxy encounters an unrecoverable internal error, such as a
memory allocation failure, which should never happen
502 when the server returns an empty, invalid or incomplete response, or
when an "rspdeny" filter blocks the response.
503 when no server was available to handle the request, or in response to
monitoring requests which match the "monitor fail" condition
504 when the response timeout strikes before the server responds
The error 4xx and 5xx codes above may be customized (see "errorloc" in section
4.2).
1.3.2. The response headers
---------------------------
Response headers work exactly like request headers, and as such, HAProxy uses
the same parsing function for both. Please refer to paragraph 1.2.2 for more
details.
2. Configuring HAProxy
----------------------
2.1. Configuration file format
------------------------------
HAProxy's configuration process involves 3 major sources of parameters :
- the arguments from the command-line, which always take precedence
- the "global" section, which sets process-wide parameters
- the proxies sections which can take form of "defaults", "listen",
"frontend" and "backend".
The configuration file syntax consists in lines beginning with a keyword
referenced in this manual, optionally followed by one or several parameters
delimited by spaces. If spaces have to be entered in strings, then they must be
preceded by a backslash ('\') to be escaped. Backslashes also have to be
escaped by doubling them.
2.2. Time format
----------------
Some parameters involve values representing time, such as timeouts. These
values are generally expressed in milliseconds (unless explicitly stated
otherwise) but may be expressed in any other unit by suffixing the unit to the
numeric value. It is important to consider this because it will not be repeated
for every keyword. Supported units are :
- us : microseconds. 1 microsecond = 1/1000000 second
- ms : milliseconds. 1 millisecond = 1/1000 second. This is the default.
- s : seconds. 1s = 1000ms
- m : minutes. 1m = 60s = 60000ms
- h : hours. 1h = 60m = 3600s = 3600000ms
- d : days. 1d = 24h = 1440m = 86400s = 86400000ms
2.3. Examples
-------------
# Simple configuration for an HTTP proxy listening on port 80 on all
# interfaces and forwarding requests to a single backend "servers" with a
# single server "server1" listening on 127.0.0.1:8000
global
daemon
maxconn 256
defaults
mode http
timeout connect 5000ms
timeout client 50000ms
timeout server 50000ms
frontend http-in
bind *:80
default_backend servers
backend servers
server server1 127.0.0.1:8000 maxconn 32
# The same configuration defined with a single listen block. Shorter but
# less expressive, especially in HTTP mode.
global
daemon
maxconn 256
defaults
mode http
timeout connect 5000ms
timeout client 50000ms
timeout server 50000ms
listen http-in
bind *:80
server server1 127.0.0.1:8000 maxconn 32
Assuming haproxy is in $PATH, test these configurations in a shell with:
$ sudo haproxy -f configuration.conf -c
3. Global parameters
--------------------
Parameters in the "global" section are process-wide and often OS-specific. They
are generally set once for all and do not need being changed once correct. Some
of them have command-line equivalents.
The following keywords are supported in the "global" section :
* Process management and security
- ca-base
- chroot
- crt-base
- daemon
- gid
- group
- log
- log-send-hostname
- nbproc
- pidfile
- uid
- ulimit-n
- user
- stats
- node
- description
- unix-bind
* Performance tuning
- maxconn
- maxconnrate
- maxcomprate
- maxcompcpuusage
- maxpipes
- maxsslconn
- noepoll
- nokqueue
- nopoll
- nosplice
- spread-checks
- tune.bufsize
- tune.chksize
- tune.comp.maxlevel
- tune.http.cookielen
- tune.http.maxhdr
- tune.maxaccept
- tune.maxpollevents
- tune.maxrewrite
- tune.pipesize
- tune.rcvbuf.client
- tune.rcvbuf.server
- tune.sndbuf.client
- tune.sndbuf.server
- tune.ssl.cachesize
- tune.ssl.lifetime
- tune.ssl.maxrecord
- tune.zlib.memlevel
- tune.zlib.windowsize
* Debugging
- debug
- quiet
3.1. Process management and security
------------------------------------
ca-base <dir>
Assigns a default directory to fetch SSL CA certificates and CRLs from when a
relative path is used with "ca-file" or "crl-file" directives. Absolute
locations specified in "ca-file" and "crl-file" prevail and ignore "ca-base".
chroot <jail dir>
Changes current directory to <jail dir> and performs a chroot() there before
dropping privileges. This increases the security level in case an unknown
vulnerability would be exploited, since it would make it very hard for the
attacker to exploit the system. This only works when the process is started
with superuser privileges. It is important to ensure that <jail_dir> is both
empty and unwritable to anyone.
cpu-map <"all"|"odd"|"even"|process_num> <cpu-set>...
On Linux 2.6 and above, it is possible to bind a process to a specific CPU
set. This means that the process will never run on other CPUs. The "cpu-map"
directive specifies CPU sets for process sets. The first argument is the
process number to bind. This process must have a number between 1 and 32,
and any process IDs above nbproc are ignored. It is possible to specify all
processes at once using "all", only odd numbers using "odd" or even numbers
using "even", just like with the "bind-process" directive. The second and
forthcoming arguments are CPU sets. Each CPU set is either a unique number
between 0 and 31 or a range with two such numbers delimited by a dash ('-').
Multiple CPU numbers or ranges may be specified, and the processes will be
allowed to bind to all of them. Obviously, multiple "cpu-map" directives may
be specified. Each "cpu-map" directive will replace the previous ones when
they overlap.
crt-base <dir>
Assigns a default directory to fetch SSL certificates from when a relative
path is used with "crtfile" directives. Absolute locations specified after
"crtfile" prevail and ignore "crt-base".
daemon
Makes the process fork into background. This is the recommended mode of
operation. It is equivalent to the command line "-D" argument. It can be
disabled by the command line "-db" argument.
gid <number>
Changes the process' group ID to <number>. It is recommended that the group
ID is dedicated to HAProxy or to a small set of similar daemons. HAProxy must
be started with a user belonging to this group, or with superuser privileges.
Note that if haproxy is started from a user having supplementary groups, it
will only be able to drop these groups if started with superuser privileges.
See also "group" and "uid".
group <group name>
Similar to "gid" but uses the GID of group name <group name> from /etc/group.
See also "gid" and "user".
log <address> <facility> [max level [min level]]
Adds a global syslog server. Up to two global servers can be defined. They
will receive logs for startups and exits, as well as all logs from proxies
configured with "log global".
<address> can be one of:
- An IPv4 address optionally followed by a colon and a UDP port. If
no port is specified, 514 is used by default (the standard syslog
port).
- An IPv6 address followed by a colon and optionally a UDP port. If
no port is specified, 514 is used by default (the standard syslog
port).
- A filesystem path to a UNIX domain socket, keeping in mind
considerations for chroot (be sure the path is accessible inside
the chroot) and uid/gid (be sure the path is appropriately
writeable).
Any part of the address string may reference any number of environment
variables by preceding their name with a dollar sign ('$') and
optionally enclosing them with braces ('{}'), similarly to what is done
in Bourne shell.
<facility> must be one of the 24 standard syslog facilities :
kern user mail daemon auth syslog lpr news
uucp cron auth2 ftp ntp audit alert cron2
local0 local1 local2 local3 local4 local5 local6 local7
An optional level can be specified to filter outgoing messages. By default,
all messages are sent. If a maximum level is specified, only messages with a
severity at least as important as this level will be sent. An optional minimum
level can be specified. If it is set, logs emitted with a more severe level
than this one will be capped to this level. This is used to avoid sending
"emerg" messages on all terminals on some default syslog configurations.
Eight levels are known :
emerg alert crit err warning notice info debug
log-send-hostname [<string>]
Sets the hostname field in the syslog header. If optional "string" parameter
is set the header is set to the string contents, otherwise uses the hostname
of the system. Generally used if one is not relaying logs through an
intermediate syslog server or for simply customizing the hostname printed in
the logs.
log-tag <string>
Sets the tag field in the syslog header to this string. It defaults to the
program name as launched from the command line, which usually is "haproxy".
Sometimes it can be useful to differentiate between multiple processes
running on the same host.
nbproc <number>
Creates <number> processes when going daemon. This requires the "daemon"
mode. By default, only one process is created, which is the recommended mode
of operation. For systems limited to small sets of file descriptors per
process, it may be needed to fork multiple daemons. USING MULTIPLE PROCESSES
IS HARDER TO DEBUG AND IS REALLY DISCOURAGED. See also "daemon".
pidfile <pidfile>
Writes pids of all daemons into file <pidfile>. This option is equivalent to
the "-p" command line argument. The file must be accessible to the user
starting the process. See also "daemon".
stats bind-process [ all | odd | even | <number 1-32>[-<number 1-32>] ] ...
Limits the stats socket to a certain set of processes numbers. By default the
stats socket is bound to all processes, causing a warning to be emitted when
nbproc is greater than 1 because there is no way to select the target process
when connecting. However, by using this setting, it becomes possible to pin
the stats socket to a specific set of processes, typically the first one. The
warning will automatically be disabled when this setting is used, whatever
the number of processes used.
stats socket [<address:port>|<path>] [param*]
Binds a UNIX socket to <path> or a TCPv4/v6 address to <address:port>.
Connections to this socket will return various statistics outputs and even
allow some commands to be issued to change some runtime settings. Please
consult section 9.2 "Unix Socket commands" for more details.
All parameters supported by "bind" lines are supported, for instance to
restrict access to some users or their access rights. Please consult
section 5.1 for more information.
stats timeout <timeout, in milliseconds>
The default timeout on the stats socket is set to 10 seconds. It is possible
to change this value with "stats timeout". The value must be passed in
milliseconds, or be suffixed by a time unit among { us, ms, s, m, h, d }.
stats maxconn <connections>
By default, the stats socket is limited to 10 concurrent connections. It is
possible to change this value with "stats maxconn".
uid <number>
Changes the process' user ID to <number>. It is recommended that the user ID
is dedicated to HAProxy or to a small set of similar daemons. HAProxy must
be started with superuser privileges in order to be able to switch to another
one. See also "gid" and "user".
ulimit-n <number>
Sets the maximum number of per-process file-descriptors to <number>. By
default, it is automatically computed, so it is recommended not to use this
option.
unix-bind [ prefix <prefix> ] [ mode <mode> ] [ user <user> ] [ uid <uid> ]
[ group <group> ] [ gid <gid> ]
Fixes common settings to UNIX listening sockets declared in "bind" statements.
This is mainly used to simplify declaration of those UNIX sockets and reduce
the risk of errors, since those settings are most commonly required but are
also process-specific. The <prefix> setting can be used to force all socket
path to be relative to that directory. This might be needed to access another
component's chroot. Note that those paths are resolved before haproxy chroots
itself, so they are absolute. The <mode>, <user>, <uid>, <group> and <gid>
all have the same meaning as their homonyms used by the "bind" statement. If
both are specified, the "bind" statement has priority, meaning that the
"unix-bind" settings may be seen as process-wide default settings.
user <user name>
Similar to "uid" but uses the UID of user name <user name> from /etc/passwd.
See also "uid" and "group".
node <name>
Only letters, digits, hyphen and underscore are allowed, like in DNS names.
This statement is useful in HA configurations where two or more processes or
servers share the same IP address. By setting a different node-name on all
nodes, it becomes easy to immediately spot what server is handling the
traffic.
description <text>
Add a text that describes the instance.
Please note that it is required to escape certain characters (# for example)
and this text is inserted into a html page so you should avoid using
"<" and ">" characters.
3.2. Performance tuning
-----------------------
maxconn <number>
Sets the maximum per-process number of concurrent connections to <number>. It
is equivalent to the command-line argument "-n". Proxies will stop accepting
connections when this limit is reached. The "ulimit-n" parameter is
automatically adjusted according to this value. See also "ulimit-n".
maxconnrate <number>
Sets the maximum per-process number of connections per second to <number>.
Proxies will stop accepting connections when this limit is reached. It can be
used to limit the global capacity regardless of each frontend capacity. It is
important to note that this can only be used as a service protection measure,
as there will not necessarily be a fair share between frontends when the
limit is reached, so it's a good idea to also limit each frontend to some
value close to its expected share. Also, lowering tune.maxaccept can improve
fairness.
maxcomprate <number>
Sets the maximum per-process input compression rate to <number> kilobytes
pers second. For each session, if the maximum is reached, the compression
level will be decreased during the session. If the maximum is reached at the
beginning of a session, the session will not compress at all. If the maximum
is not reached, the compression level will be increased up to
tune.comp.maxlevel. A value of zero means there is no limit, this is the
default value.
maxcompcpuusage <number>
Sets the maximum CPU usage HAProxy can reach before stopping the compression
for new requests or decreasing the compression level of current requests.
It works like 'maxcomprate' but measures CPU usage instead of incoming data
bandwidth. The value is expressed in percent of the CPU used by haproxy. In
case of multiple processes (nbproc > 1), each process manages its individual
usage. A value of 100 disable the limit. The default value is 100. Setting
a lower value will prevent the compression work from slowing the whole
process down and from introducing high latencies.
maxpipes <number>
Sets the maximum per-process number of pipes to <number>. Currently, pipes
are only used by kernel-based tcp splicing. Since a pipe contains two file
descriptors, the "ulimit-n" value will be increased accordingly. The default
value is maxconn/4, which seems to be more than enough for most heavy usages.
The splice code dynamically allocates and releases pipes, and can fall back
to standard copy, so setting this value too low may only impact performance.
maxsslconn <number>
Sets the maximum per-process number of concurrent SSL connections to
<number>. By default there is no SSL-specific limit, which means that the
global maxconn setting will apply to all connections. Setting this limit
avoids having openssl use too much memory and crash when malloc returns NULL
(since it unfortunately does not reliably check for such conditions). Note
that the limit applies both to incoming and outgoing connections, so one
connection which is deciphered then ciphered accounts for 2 SSL connections.
maxzlibmem <number>
Sets the maximum amount of RAM in megabytes per process usable by the zlib.
When the maximum amount is reached, future sessions will not compress as long
as RAM is unavailable. When sets to 0, there is no limit.
The default value is 0. The value is available in bytes on the UNIX socket
with "show info" on the line "MaxZlibMemUsage", the memory used by zlib is
"ZlibMemUsage" in bytes.
noepoll
Disables the use of the "epoll" event polling system on Linux. It is
equivalent to the command-line argument "-de". The next polling system
used will generally be "poll". See also "nopoll".
nokqueue
Disables the use of the "kqueue" event polling system on BSD. It is
equivalent to the command-line argument "-dk". The next polling system
used will generally be "poll". See also "nopoll".
nopoll
Disables the use of the "poll" event polling system. It is equivalent to the
command-line argument "-dp". The next polling system used will be "select".
It should never be needed to disable "poll" since it's available on all
platforms supported by HAProxy. See also "nokqueue" and "noepoll".
nosplice
Disables the use of kernel tcp splicing between sockets on Linux. It is
equivalent to the command line argument "-dS". Data will then be copied
using conventional and more portable recv/send calls. Kernel tcp splicing is
limited to some very recent instances of kernel 2.6. Most versions between
2.6.25 and 2.6.28 are buggy and will forward corrupted data, so they must not
be used. This option makes it easier to globally disable kernel splicing in
case of doubt. See also "option splice-auto", "option splice-request" and
"option splice-response".
spread-checks <0..50, in percent>
Sometimes it is desirable to avoid sending agent and health checks to
servers at exact intervals, for instance when many logical servers are
located on the same physical server. With the help of this parameter, it
becomes possible to add some randomness in the check interval between 0
and +/- 50%. A value between 2 and 5 seems to show good results. The
default value remains at 0.
tune.bufsize <number>
Sets the buffer size to this size (in bytes). Lower values allow more
sessions to coexist in the same amount of RAM, and higher values allow some
applications with very large cookies to work. The default value is 16384 and
can be changed at build time. It is strongly recommended not to change this
from the default value, as very low values will break some services such as
statistics, and values larger than default size will increase memory usage,
possibly causing the system to run out of memory. At least the global maxconn
parameter should be decreased by the same factor as this one is increased.
If HTTP request is larger than (tune.bufsize - tune.maxrewrite), haproxy will
return HTTP 400 (Bad Request) error. Similarly if an HTTP response is larger
than this size, haproxy will return HTTP 502 (Bad Gateway).
tune.chksize <number>
Sets the check buffer size to this size (in bytes). Higher values may help
find string or regex patterns in very large pages, though doing so may imply
more memory and CPU usage. The default value is 16384 and can be changed at
build time. It is not recommended to change this value, but to use better
checks whenever possible.
tune.comp.maxlevel <number>
Sets the maximum compression level. The compression level affects CPU
usage during compression. This value affects CPU usage during compression.
Each session using compression initializes the compression algorithm with
this value. The default value is 1.
tune.http.cookielen <number>
Sets the maximum length of captured cookies. This is the maximum value that
the "capture cookie xxx len yyy" will be allowed to take, and any upper value
will automatically be truncated to this one. It is important not to set too
high a value because all cookie captures still allocate this size whatever
their configured value (they share a same pool). This value is per request
per response, so the memory allocated is twice this value per connection.
When not specified, the limit is set to 63 characters. It is recommended not
to change this value.
tune.http.maxhdr <number>
Sets the maximum number of headers in a request. When a request comes with a
number of headers greater than this value (including the first line), it is
rejected with a "400 Bad Request" status code. Similarly, too large responses
are blocked with "502 Bad Gateway". The default value is 101, which is enough
for all usages, considering that the widely deployed Apache server uses the
same limit. It can be useful to push this limit further to temporarily allow
a buggy application to work by the time it gets fixed. Keep in mind that each
new header consumes 32bits of memory for each session, so don't push this
limit too high.
tune.maxaccept <number>
Sets the maximum number of consecutive connections a process may accept in a
row before switching to other work. In single process mode, higher numbers
give better performance at high connection rates. However in multi-process
modes, keeping a bit of fairness between processes generally is better to
increase performance. This value applies individually to each listener, so
that the number of processes a listener is bound to is taken into account.
This value defaults to 64. In multi-process mode, it is divided by twice
the number of processes the listener is bound to. Setting this value to -1
completely disables the limitation. It should normally not be needed to tweak
this value.
tune.maxpollevents <number>
Sets the maximum amount of events that can be processed at once in a call to
the polling system. The default value is adapted to the operating system. It
has been noticed that reducing it below 200 tends to slightly decrease
latency at the expense of network bandwidth, and increasing it above 200
tends to trade latency for slightly increased bandwidth.
tune.maxrewrite <number>
Sets the reserved buffer space to this size in bytes. The reserved space is
used for header rewriting or appending. The first reads on sockets will never
fill more than bufsize-maxrewrite. Historically it has defaulted to half of
bufsize, though that does not make much sense since there are rarely large
numbers of headers to add. Setting it too high prevents processing of large
requests or responses. Setting it too low prevents addition of new headers
to already large requests or to POST requests. It is generally wise to set it
to about 1024. It is automatically readjusted to half of bufsize if it is
larger than that. This means you don't have to worry about it when changing
bufsize.
tune.pipesize <number>
Sets the kernel pipe buffer size to this size (in bytes). By default, pipes
are the default size for the system. But sometimes when using TCP splicing,
it can improve performance to increase pipe sizes, especially if it is
suspected that pipes are not filled and that many calls to splice() are
performed. This has an impact on the kernel's memory footprint, so this must
not be changed if impacts are not understood.
tune.rcvbuf.client <number>
tune.rcvbuf.server <number>
Forces the kernel socket receive buffer size on the client or the server side
to the specified value in bytes. This value applies to all TCP/HTTP frontends
and backends. It should normally never be set, and the default size (0) lets
the kernel autotune this value depending on the amount of available memory.
However it can sometimes help to set it to very low values (eg: 4096) in
order to save kernel memory by preventing it from buffering too large amounts
of received data. Lower values will significantly increase CPU usage though.
tune.sndbuf.client <number>
tune.sndbuf.server <number>
Forces the kernel socket send buffer size on the client or the server side to
the specified value in bytes. This value applies to all TCP/HTTP frontends
and backends. It should normally never be set, and the default size (0) lets
the kernel autotune this value depending on the amount of available memory.
However it can sometimes help to set it to very low values (eg: 4096) in
order to save kernel memory by preventing it from buffering too large amounts
of received data. Lower values will significantly increase CPU usage though.
Another use case is to prevent write timeouts with extremely slow clients due
to the kernel waiting for a large part of the buffer to be read before
notifying haproxy again.
tune.ssl.cachesize <number>
Sets the size of the global SSL session cache, in a number of blocks. A block
is large enough to contain an encoded session without peer certificate.
An encoded session with peer certificate is stored in multiple blocks
depending on the size of the peer certificate. A block use approximatively
200 bytes of memory. The default value may be forced at build time, otherwise
defaults to 20000. When the cache is full, the most idle entries are purged
and reassigned. Higher values reduce the occurrence of such a purge, hence
the number of CPU-intensive SSL handshakes by ensuring that all users keep
their session as long as possible. All entries are pre-allocated upon startup
and are shared between all processes if "nbproc" is greater than 1. Setting
this value to 0 disables the SSL session cache.
tune.ssl.lifetime <timeout>
Sets how long a cached SSL session may remain valid. This time is expressed
in seconds and defaults to 300 (5 mn). It is important to understand that it
does not guarantee that sessions will last that long, because if the cache is
full, the longest idle sessions will be purged despite their configured
lifetime. The real usefulness of this setting is to prevent sessions from
being used for too long.
tune.ssl.maxrecord <number>
Sets the maximum amount of bytes passed to SSL_write() at a time. Default
value 0 means there is no limit. Over SSL/TLS, the client can decipher the
data only once it has received a full record. With large records, it means
that clients might have to download up to 16kB of data before starting to
process them. Limiting the value can improve page load times on browsers
located over high latency or low bandwidth networks. It is suggested to find
optimal values which fit into 1 or 2 TCP segments (generally 1448 bytes over
Ethernet with TCP timestamps enabled, or 1460 when timestamps are disabled),
keeping in mind that SSL/TLS add some overhead. Typical values of 1419 and
2859 gave good results during tests. Use "strace -e trace=write" to find the
best value.
tune.zlib.memlevel <number>
Sets the memLevel parameter in zlib initialization for each session. It
defines how much memory should be allocated for the intenal compression
state. A value of 1 uses minimum memory but is slow and reduces compression
ratio, a value of 9 uses maximum memory for optimal speed. Can be a value
between 1 and 9. The default value is 8.
tune.zlib.windowsize <number>
Sets the window size (the size of the history buffer) as a parameter of the
zlib initialization for each session. Larger values of this parameter result
in better compression at the expense of memory usage. Can be a value between
8 and 15. The default value is 15.
3.3. Debugging
--------------
debug
Enables debug mode which dumps to stdout all exchanges, and disables forking
into background. It is the equivalent of the command-line argument "-d". It
should never be used in a production configuration since it may prevent full
system startup.
quiet
Do not display any message during startup. It is equivalent to the command-
line argument "-q".
3.4. Userlists
--------------
It is possible to control access to frontend/backend/listen sections or to
http stats by allowing only authenticated and authorized users. To do this,
it is required to create at least one userlist and to define users.
userlist <listname>
Creates new userlist with name <listname>. Many independent userlists can be
used to store authentication & authorization data for independent customers.
group <groupname> [users <user>,<user>,(...)]
Adds group <groupname> to the current userlist. It is also possible to
attach users to this group by using a comma separated list of names
proceeded by "users" keyword.
user <username> [password|insecure-password <password>]
[groups <group>,<group>,(...)]
Adds user <username> to the current userlist. Both secure (encrypted) and
insecure (unencrypted) passwords can be used. Encrypted passwords are
evaluated using the crypt(3) function so depending of the system's
capabilities, different algorithms are supported. For example modern Glibc
based Linux system supports MD5, SHA-256, SHA-512 and of course classic,
DES-based method of crypting passwords.
Example:
userlist L1
group G1 users tiger,scott
group G2 users xdb,scott
user tiger password $6$k6y3o.eP$JlKBx9za9667qe4(...)xHSwRv6J.C0/D7cV91
user scott insecure-password elgato
user xdb insecure-password hello
userlist L2
group G1
group G2
user tiger password $6$k6y3o.eP$JlKBx(...)xHSwRv6J.C0/D7cV91 groups G1
user scott insecure-password elgato groups G1,G2
user xdb insecure-password hello groups G2
Please note that both lists are functionally identical.
3.5. Peers
----------
It is possible to synchronize server entries in stick tables between several
haproxy instances over TCP connections in a multi-master fashion. Each instance
pushes its local updates and insertions to remote peers. Server IDs are used to
identify servers remotely, so it is important that configurations look similar
or at least that the same IDs are forced on each server on all participants.
Interrupted exchanges are automatically detected and recovered from the last
known point. In addition, during a soft restart, the old process connects to
the new one using such a TCP connection to push all its entries before the new
process tries to connect to other peers. That ensures very fast replication
during a reload, it typically takes a fraction of a second even for large
tables.
peers <peersect>
Creates a new peer list with name <peersect>. It is an independent section,
which is referenced by one or more stick-tables.
peer <peername> <ip>:<port>
Defines a peer inside a peers section.
If <peername> is set to the local peer name (by default hostname, or forced
using "-L" command line option), haproxy will listen for incoming remote peer
connection on <ip>:<port>. Otherwise, <ip>:<port> defines where to connect to
to join the remote peer, and <peername> is used at the protocol level to
identify and validate the remote peer on the server side.
During a soft restart, local peer <ip>:<port> is used by the old instance to
connect the new one and initiate a complete replication (teaching process).
It is strongly recommended to have the exact same peers declaration on all
peers and to only rely on the "-L" command line argument to change the local
peer name. This makes it easier to maintain coherent configuration files
across all peers.
Any part of the address string may reference any number of environment
variables by preceding their name with a dollar sign ('$') and optionally
enclosing them with braces ('{}'), similarly to what is done in Bourne shell.
Example:
peers mypeers
peer haproxy1 192.168.0.1:1024
peer haproxy2 192.168.0.2:1024
peer haproxy3 10.2.0.1:1024
backend mybackend
mode tcp
balance roundrobin
stick-table type ip size 20k peers mypeers
stick on src
server srv1 192.168.0.30:80
server srv2 192.168.0.31:80
4. Proxies
----------
Proxy configuration can be located in a set of sections :
- defaults <name>
- frontend <name>
- backend <name>
- listen <name>
A "defaults" section sets default parameters for all other sections following
its declaration. Those default parameters are reset by the next "defaults"
section. See below for the list of parameters which can be set in a "defaults"
section. The name is optional but its use is encouraged for better readability.
A "frontend" section describes a set of listening sockets accepting client
connections.
A "backend" section describes a set of servers to which the proxy will connect
to forward incoming connections.
A "listen" section defines a complete proxy with its frontend and backend
parts combined in one section. It is generally useful for TCP-only traffic.
All proxy names must be formed from upper and lower case letters, digits,
'-' (dash), '_' (underscore) , '.' (dot) and ':' (colon). ACL names are
case-sensitive, which means that "www" and "WWW" are two different proxies.
Historically, all proxy names could overlap, it just caused troubles in the
logs. Since the introduction of content switching, it is mandatory that two
proxies with overlapping capabilities (frontend/backend) have different names.
However, it is still permitted that a frontend and a backend share the same
name, as this configuration seems to be commonly encountered.
Right now, two major proxy modes are supported : "tcp", also known as layer 4,
and "http", also known as layer 7. In layer 4 mode, HAProxy simply forwards
bidirectional traffic between two sides. In layer 7 mode, HAProxy analyzes the
protocol, and can interact with it by allowing, blocking, switching, adding,
modifying, or removing arbitrary contents in requests or responses, based on
arbitrary criteria.
4.1. Proxy keywords matrix
--------------------------
The following list of keywords is supported. Most of them may only be used in a
limited set of section types. Some of them are marked as "deprecated" because
they are inherited from an old syntax which may be confusing or functionally
limited, and there are new recommended keywords to replace them. Keywords
marked with "(*)" can be optionally inverted using the "no" prefix, eg. "no
option contstats". This makes sense when the option has been enabled by default
and must be disabled for a specific instance. Such options may also be prefixed
with "default" in order to restore default settings regardless of what has been
specified in a previous "defaults" section.
keyword defaults frontend listen backend
------------------------------------+----------+----------+---------+---------
acl - X X X
appsession - - X X
backlog X X X -
balance X - X X
bind - X X -
bind-process X X X X
block - X X X
capture cookie - X X -
capture request header - X X -
capture response header - X X -
clitimeout (deprecated) X X X -
compression X X X X
contimeout (deprecated) X - X X
cookie X - X X
default-server X - X X
default_backend X X X -
description - X X X
disabled X X X X
dispatch - - X X
enabled X X X X
errorfile X X X X
errorloc X X X X
errorloc302 X X X X
-- keyword -------------------------- defaults - frontend - listen -- backend -
errorloc303 X X X X
force-persist - X X X
fullconn X - X X
grace X X X X
hash-type X - X X
http-check disable-on-404 X - X X
http-check expect - - X X
http-check send-state X - X X
http-request - X X X
http-response - X X X
tcp-check expect - - X X
tcp-check send - - X X
tcp-check send-binary - - X X
http-send-name-header - - X X
id - X X X
ignore-persist - X X X
log (*) X X X X
maxconn X X X -
mode X X X X
monitor fail - X X -
monitor-net X X X -
monitor-uri X X X -
option abortonclose (*) X - X X
option accept-invalid-http-request (*) X X X -
option accept-invalid-http-response (*) X - X X
option allbackups (*) X - X X
option checkcache (*) X - X X
option clitcpka (*) X X X -
option contstats (*) X X X -
option dontlog-normal (*) X X X -
option dontlognull (*) X X X -
option forceclose (*) X X X X
-- keyword -------------------------- defaults - frontend - listen -- backend -
option forwardfor X X X X
option http-keep-alive (*) X X X X
option http-no-delay (*) X X X X
option http-pretend-keepalive (*) X X X X
option http-server-close (*) X X X X
option http-use-proxy-header (*) X X X -
option httpchk X - X X
option httpclose (*) X X X X
option httplog X X X X
option http_proxy (*) X X X X
option independent-streams (*) X X X X
option ldap-check X - X X
option log-health-checks (*) X - X X
option log-separate-errors (*) X X X -
option logasap (*) X X X -
option mysql-check X - X X
option pgsql-check X - X X
option nolinger (*) X X X X
option originalto X X X X
option persist (*) X - X X
option redispatch (*) X - X X
option redis-check X - X X
option smtpchk X - X X
option socket-stats (*) X X X -
option splice-auto (*) X X X X
option splice-request (*) X X X X
option splice-response (*) X X X X
option srvtcpka (*) X - X X
option ssl-hello-chk X - X X
-- keyword -------------------------- defaults - frontend - listen -- backend -
option tcp-check X - X X
option tcp-smart-accept (*) X X X -
option tcp-smart-connect (*) X - X X
option tcpka X X X X
option tcplog X X X X
option transparent (*) X - X X
persist rdp-cookie X - X X
rate-limit sessions X X X -
redirect - X X X
redisp (deprecated) X - X X
redispatch (deprecated) X - X X
reqadd - X X X
reqallow - X X X
reqdel - X X X
reqdeny - X X X
reqiallow - X X X
reqidel - X X X
reqideny - X X X
reqipass - X X X
reqirep - X X X
reqisetbe - X X X
reqitarpit - X X X
reqpass - X X X
reqrep - X X X
-- keyword -------------------------- defaults - frontend - listen -- backend -
reqsetbe - X X X
reqtarpit - X X X
retries X - X X
rspadd - X X X
rspdel - X X X
rspdeny - X X X
rspidel - X X X
rspideny - X X X
rspirep - X X X
rsprep - X X X
server - - X X
source X - X X
srvtimeout (deprecated) X - X X
stats admin - - X X
stats auth X - X X
stats enable X - X X
stats hide-version X - X X
stats http-request - - X X
stats realm X - X X
stats refresh X - X X
stats scope X - X X
stats show-desc X - X X
stats show-legends X - X X
stats show-node X - X X
stats uri X - X X
-- keyword -------------------------- defaults - frontend - listen -- backend -
stick match - - X X
stick on - - X X
stick store-request - - X X
stick store-response - - X X
stick-table - - X X
tcp-request connection - X X -
tcp-request content - X X X
tcp-request inspect-delay - X X X
tcp-response content - - X X
tcp-response inspect-delay - - X X
timeout check X - X X
timeout client X X X -
timeout clitimeout (deprecated) X X X -
timeout connect X - X X
timeout contimeout (deprecated) X - X X
timeout http-keep-alive X X X X
timeout http-request X X X X
timeout queue X - X X
timeout server X - X X
timeout srvtimeout (deprecated) X - X X
timeout tarpit X X X X
timeout tunnel X - X X
transparent (deprecated) X - X X
unique-id-format X X X -
unique-id-header X X X -
use_backend - X X -
use-server - - X X
------------------------------------+----------+----------+---------+---------
keyword defaults frontend listen backend
4.2. Alphabetically sorted keywords reference
---------------------------------------------
This section provides a description of each keyword and its usage.
acl <aclname> <criterion> [flags] [operator] <value> ...
Declare or complete an access list.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | yes
Example:
acl invalid_src src 0.0.0.0/7 224.0.0.0/3
acl invalid_src src_port 0:1023
acl local_dst hdr(host) -i localhost
See section 7 about ACL usage.
appsession <cookie> len <length> timeout <holdtime>
[request-learn] [prefix] [mode <path-parameters|query-string>]
Define session stickiness on an existing application cookie.
May be used in sections : defaults | frontend | listen | backend
no | no | yes | yes
Arguments :
<cookie> this is the name of the cookie used by the application and which
HAProxy will have to learn for each new session.
<length> this is the max number of characters that will be memorized and
checked in each cookie value.
<holdtime> this is the time after which the cookie will be removed from
memory if unused. If no unit is specified, this time is in
milliseconds.
request-learn
If this option is specified, then haproxy will be able to learn
the cookie found in the request in case the server does not
specify any in response. This is typically what happens with
PHPSESSID cookies, or when haproxy's session expires before
the application's session and the correct server is selected.
It is recommended to specify this option to improve reliability.
prefix When this option is specified, haproxy will match on the cookie
prefix (or URL parameter prefix). The appsession value is the
data following this prefix.
Example :
appsession ASPSESSIONID len 64 timeout 3h prefix
This will match the cookie ASPSESSIONIDXXXX=XXXXX,
the appsession value will be XXXX=XXXXX.
mode This option allows to change the URL parser mode.
2 modes are currently supported :
- path-parameters :
The parser looks for the appsession in the path parameters
part (each parameter is separated by a semi-colon), which is
convenient for JSESSIONID for example.
This is the default mode if the option is not set.
- query-string :
In this mode, the parser will look for the appsession in the
query string.
When an application cookie is defined in a backend, HAProxy will check when
the server sets such a cookie, and will store its value in a table, and
associate it with the server's identifier. Up to <length> characters from
the value will be retained. On each connection, haproxy will look for this
cookie both in the "Cookie:" headers, and as a URL parameter (depending on
the mode used). If a known value is found, the client will be directed to the
server associated with this value. Otherwise, the load balancing algorithm is
applied. Cookies are automatically removed from memory when they have been
unused for a duration longer than <holdtime>.
The definition of an application cookie is limited to one per backend.
Note : Consider not using this feature in multi-process mode (nbproc > 1)
unless you know what you do : memory is not shared between the
processes, which can result in random behaviours.
Example :
appsession JSESSIONID len 52 timeout 3h
See also : "cookie", "capture cookie", "balance", "stick", "stick-table",
"ignore-persist", "nbproc" and "bind-process".
backlog <conns>
Give hints to the system about the approximate listen backlog desired size
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<conns> is the number of pending connections. Depending on the operating
system, it may represent the number of already acknowledged
connections, of non-acknowledged ones, or both.
In order to protect against SYN flood attacks, one solution is to increase
the system's SYN backlog size. Depending on the system, sometimes it is just
tunable via a system parameter, sometimes it is not adjustable at all, and
sometimes the system relies on hints given by the application at the time of
the listen() syscall. By default, HAProxy passes the frontend's maxconn value
to the listen() syscall. On systems which can make use of this value, it can
sometimes be useful to be able to specify a different value, hence this
backlog parameter.
On Linux 2.4, the parameter is ignored by the system. On Linux 2.6, it is
used as a hint and the system accepts up to the smallest greater power of
two, and never more than some limits (usually 32768).
See also : "maxconn" and the target operating system's tuning guide.
balance <algorithm> [ <arguments> ]
balance url_param <param> [check_post [<max_wait>]]
Define the load balancing algorithm to be used in a backend.
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<algorithm> is the algorithm used to select a server when doing load
balancing. This only applies when no persistence information
is available, or when a connection is redispatched to another
server. <algorithm> may be one of the following :
roundrobin Each server is used in turns, according to their weights.
This is the smoothest and fairest algorithm when the server's
processing time remains equally distributed. This algorithm
is dynamic, which means that server weights may be adjusted
on the fly for slow starts for instance. It is limited by
design to 4095 active servers per backend. Note that in some
large farms, when a server becomes up after having been down
for a very short time, it may sometimes take a few hundreds
requests for it to be re-integrated into the farm and start
receiving traffic. This is normal, though very rare. It is
indicated here in case you would have the chance to observe
it, so that you don't worry.
static-rr Each server is used in turns, according to their weights.
This algorithm is as similar to roundrobin except that it is
static, which means that changing a server's weight on the
fly will have no effect. On the other hand, it has no design
limitation on the number of servers, and when a server goes
up, it is always immediately reintroduced into the farm, once
the full map is recomputed. It also uses slightly less CPU to
run (around -1%).
leastconn The server with the lowest number of connections receives the
connection. Round-robin is performed within groups of servers
of the same load to ensure that all servers will be used. Use
of this algorithm is recommended where very long sessions are
expected, such as LDAP, SQL, TSE, etc... but is not very well
suited for protocols using short sessions such as HTTP. This
algorithm is dynamic, which means that server weights may be
adjusted on the fly for slow starts for instance.
first The first server with available connection slots receives the
connection. The servers are choosen from the lowest numeric
identifier to the highest (see server parameter "id"), which
defaults to the server's position in the farm. Once a server
reaches its maxconn value, the next server is used. It does
not make sense to use this algorithm without setting maxconn.
The purpose of this algorithm is to always use the smallest
number of servers so that extra servers can be powered off
during non-intensive hours. This algorithm ignores the server
weight, and brings more benefit to long session such as RDP
or IMAP than HTTP, though it can be useful there too. In
order to use this algorithm efficiently, it is recommended
that a cloud controller regularly checks server usage to turn
them off when unused, and regularly checks backend queue to
turn new servers on when the queue inflates. Alternatively,
using "http-check send-state" may inform servers on the load.
source The source IP address is hashed and divided by the total
weight of the running servers to designate which server will
receive the request. This ensures that the same client IP
address will always reach the same server as long as no
server goes down or up. If the hash result changes due to the
number of running servers changing, many clients will be
directed to a different server. This algorithm is generally
used in TCP mode where no cookie may be inserted. It may also
be used on the Internet to provide a best-effort stickiness
to clients which refuse session cookies. This algorithm is
static by default, which means that changing a server's
weight on the fly will have no effect, but this can be
changed using "hash-type".
uri This algorithm hashes either the left part of the URI (before
the question mark) or the whole URI (if the "whole" parameter
is present) and divides the hash value by the total weight of
the running servers. The result designates which server will
receive the request. This ensures that the same URI will
always be directed to the same server as long as no server
goes up or down. This is used with proxy caches and
anti-virus proxies in order to maximize the cache hit rate.
Note that this algorithm may only be used in an HTTP backend.
This algorithm is static by default, which means that
changing a server's weight on the fly will have no effect,
but this can be changed using "hash-type".
This algorithm supports two optional parameters "len" and
"depth", both followed by a positive integer number. These
options may be helpful when it is needed to balance servers
based on the beginning of the URI only. The "len" parameter
indicates that the algorithm should only consider that many
characters at the beginning of the URI to compute the hash.
Note that having "len" set to 1 rarely makes sense since most
URIs start with a leading "/".
The "depth" parameter indicates the maximum directory depth
to be used to compute the hash. One level is counted for each
slash in the request. If both parameters are specified, the
evaluation stops when either is reached.
url_param The URL parameter specified in argument will be looked up in
the query string of each HTTP GET request.
If the modifier "check_post" is used, then an HTTP POST
request entity will be searched for the parameter argument,
when it is not found in a query string after a question mark
('?') in the URL. Optionally, specify a number of octets to
wait for before attempting to search the message body. If the
entity can not be searched, then round robin is used for each
request. For instance, if your clients always send the LB
parameter in the first 128 bytes, then specify that. The
default is 48. The entity data will not be scanned until the
required number of octets have arrived at the gateway, this
is the minimum of: (default/max_wait, Content-Length or first
chunk length). If Content-Length is missing or zero, it does
not need to wait for more data than the client promised to
send. When Content-Length is present and larger than
<max_wait>, then waiting is limited to <max_wait> and it is
assumed that this will be enough data to search for the
presence of the parameter. In the unlikely event that
Transfer-Encoding: chunked is used, only the first chunk is
scanned. Parameter values separated by a chunk boundary, may
be randomly balanced if at all.
If the parameter is found followed by an equal sign ('=') and
a value, then the value is hashed and divided by the total
weight of the running servers. The result designates which
server will receive the request.
This is used to track user identifiers in requests and ensure
that a same user ID will always be sent to the same server as
long as no server goes up or down. If no value is found or if
the parameter is not found, then a round robin algorithm is
applied. Note that this algorithm may only be used in an HTTP
backend. This algorithm is static by default, which means
that changing a server's weight on the fly will have no
effect, but this can be changed using "hash-type".
hdr(<name>) The HTTP header <name> will be looked up in each HTTP
request. Just as with the equivalent ACL 'hdr()' function,
the header name in parenthesis is not case sensitive. If the
header is absent or if it does not contain any value, the
roundrobin algorithm is applied instead.
An optional 'use_domain_only' parameter is available, for
reducing the hash algorithm to the main domain part with some
specific headers such as 'Host'. For instance, in the Host
value "haproxy.1wt.eu", only "1wt" will be considered.
This algorithm is static by default, which means that
changing a server's weight on the fly will have no effect,
but this can be changed using "hash-type".
rdp-cookie
rdp-cookie(<name>)
The RDP cookie <name> (or "mstshash" if omitted) will be
looked up and hashed for each incoming TCP request. Just as
with the equivalent ACL 'req_rdp_cookie()' function, the name
is not case-sensitive. This mechanism is useful as a degraded
persistence mode, as it makes it possible to always send the
same user (or the same session ID) to the same server. If the
cookie is not found, the normal roundrobin algorithm is
used instead.
Note that for this to work, the frontend must ensure that an
RDP cookie is already present in the request buffer. For this
you must use 'tcp-request content accept' rule combined with
a 'req_rdp_cookie_cnt' ACL.
This algorithm is static by default, which means that
changing a server's weight on the fly will have no effect,
but this can be changed using "hash-type".
See also the rdp_cookie pattern fetch function.
<arguments> is an optional list of arguments which may be needed by some
algorithms. Right now, only "url_param" and "uri" support an
optional argument.
balance uri [len <len>] [depth <depth>]
balance url_param <param> [check_post [<max_wait>]]
The load balancing algorithm of a backend is set to roundrobin when no other
algorithm, mode nor option have been set. The algorithm may only be set once
for each backend.
Examples :
balance roundrobin
balance url_param userid
balance url_param session_id check_post 64
balance hdr(User-Agent)
balance hdr(host)
balance hdr(Host) use_domain_only
Note: the following caveats and limitations on using the "check_post"
extension with "url_param" must be considered :
- all POST requests are eligible for consideration, because there is no way
to determine if the parameters will be found in the body or entity which
may contain binary data. Therefore another method may be required to
restrict consideration of POST requests that have no URL parameters in
the body. (see acl reqideny http_end)
- using a <max_wait> value larger than the request buffer size does not
make sense and is useless. The buffer size is set at build time, and
defaults to 16 kB.
- Content-Encoding is not supported, the parameter search will probably
fail; and load balancing will fall back to Round Robin.
- Expect: 100-continue is not supported, load balancing will fall back to
Round Robin.
- Transfer-Encoding (RFC2616 3.6.1) is only supported in the first chunk.
If the entire parameter value is not present in the first chunk, the
selection of server is undefined (actually, defined by how little
actually appeared in the first chunk).
- This feature does not support generation of a 100, 411 or 501 response.
- In some cases, requesting "check_post" MAY attempt to scan the entire
contents of a message body. Scanning normally terminates when linear
white space or control characters are found, indicating the end of what
might be a URL parameter list. This is probably not a concern with SGML
type message bodies.
See also : "dispatch", "cookie", "appsession", "transparent", "hash-type" and
"http_proxy".
bind [<address>]:<port_range> [, ...] [param*]
bind /<path> [, ...] [param*]
Define one or several listening addresses and/or ports in a frontend.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | no
Arguments :
<address> is optional and can be a host name, an IPv4 address, an IPv6
address, or '*'. It designates the address the frontend will
listen on. If unset, all IPv4 addresses of the system will be
listened on. The same will apply for '*' or the system's
special address "0.0.0.0". The IPv6 equivalent is '::'.
Optionally, an address family prefix may be used before the
address to force the family regardless of the address format,
which can be useful to specify a path to a unix socket with
no slash ('/'). Currently supported prefixes are :
- 'ipv4@' -> address is always IPv4
- 'ipv6@' -> address is always IPv6
- 'unix@' -> address is a path to a local unix socket
- 'fd@<n>' -> use file descriptor <n> inherited from the
parent. The fd must be bound and may or may not already
be listening.
Any part of the address string may reference any number of
environment variables by preceding their name with a dollar
sign ('$') and optionally enclosing them with braces ('{}'),
similarly to what is done in Bourne shell.
<port_range> is either a unique TCP port, or a port range for which the
proxy will accept connections for the IP address specified
above. The port is mandatory for TCP listeners. Note that in
the case of an IPv6 address, the port is always the number
after the last colon (':'). A range can either be :
- a numerical port (ex: '80')
- a dash-delimited ports range explicitly stating the lower
and upper bounds (ex: '2000-2100') which are included in
the range.
Particular care must be taken against port ranges, because
every <address:port> couple consumes one socket (= a file
descriptor), so it's easy to consume lots of descriptors
with a simple range, and to run out of sockets. Also, each
<address:port> couple must be used only once among all
instances running on a same system. Please note that binding
to ports lower than 1024 generally require particular
privileges to start the program, which are independent of
the 'uid' parameter.
<path> is a UNIX socket path beginning with a slash ('/'). This is
alternative to the TCP listening port. Haproxy will then
receive UNIX connections on the socket located at this place.
The path must begin with a slash and by default is absolute.
It can be relative to the prefix defined by "unix-bind" in
the global section. Note that the total length of the prefix
followed by the socket path cannot exceed some system limits
for UNIX sockets, which commonly are set to 107 characters.
<param*> is a list of parameters common to all sockets declared on the
same line. These numerous parameters depend on OS and build
options and have a complete section dedicated to them. Please
refer to section 5 to for more details.
It is possible to specify a list of address:port combinations delimited by
commas. The frontend will then listen on all of these addresses. There is no
fixed limit to the number of addresses and ports which can be listened on in
a frontend, as well as there is no limit to the number of "bind" statements
in a frontend.
Example :
listen http_proxy
bind :80,:443
bind 10.0.0.1:10080,10.0.0.1:10443
bind /var/run/ssl-frontend.sock user root mode 600 accept-proxy
listen http_https_proxy
bind :80
bind :443 ssl crt /etc/haproxy/site.pem
listen http_https_proxy_explicit
bind ipv6@:80
bind ipv4@public_ssl:443 ssl crt /etc/haproxy/site.pem
bind unix@ssl-frontend.sock user root mode 600 accept-proxy
listen external_bind_app1
bind fd@${FD_APP1}
See also : "source", "option forwardfor", "unix-bind" and the PROXY protocol
documentation, and section 5 about bind options.
bind-process [ all | odd | even | <number 1-32>[-<number 1-32>] ] ...
Limit visibility of an instance to a certain set of processes numbers.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
all All process will see this instance. This is the default. It
may be used to override a default value.
odd This instance will be enabled on processes 1,3,5,...31. This
option may be combined with other numbers.
even This instance will be enabled on processes 2,4,6,...32. This
option may be combined with other numbers. Do not use it
with less than 2 processes otherwise some instances might be
missing from all processes.
number The instance will be enabled on this process number or range,
whose values must all be between 1 and 32. You must be
careful not to reference a process number greater than the
configured global.nbproc, otherwise some instances might be
missing from all processes.
This keyword limits binding of certain instances to certain processes. This
is useful in order not to have too many processes listening to the same
ports. For instance, on a dual-core machine, it might make sense to set
'nbproc 2' in the global section, then distributes the listeners among 'odd'
and 'even' instances.
At the moment, it is not possible to reference more than 32 processes using
this keyword, but this should be more than enough for most setups. Please
note that 'all' really means all processes and is not limited to the first
32.
If some backends are referenced by frontends bound to other processes, the
backend automatically inherits the frontend's processes.
Example :
listen app_ip1
bind 10.0.0.1:80
bind-process odd
listen app_ip2
bind 10.0.0.2:80
bind-process even
listen management
bind 10.0.0.3:80
bind-process 1 2 3 4
listen management
bind 10.0.0.4:80
bind-process 1-4
See also : "nbproc" in global section.
block { if | unless } <condition>
Block a layer 7 request if/unless a condition is matched
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | yes
The HTTP request will be blocked very early in the layer 7 processing
if/unless <condition> is matched. A 403 error will be returned if the request
is blocked. The condition has to reference ACLs (see section 7). This is
typically used to deny access to certain sensitive resources if some
conditions are met or not met. There is no fixed limit to the number of
"block" statements per instance.
Example:
acl invalid_src src 0.0.0.0/7 224.0.0.0/3
acl invalid_src src_port 0:1023
acl local_dst hdr(host) -i localhost
block if invalid_src || local_dst
See section 7 about ACL usage.
capture cookie <name> len <length>
Capture and log a cookie in the request and in the response.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | no
Arguments :
<name> is the beginning of the name of the cookie to capture. In order
to match the exact name, simply suffix the name with an equal
sign ('='). The full name will appear in the logs, which is
useful with application servers which adjust both the cookie name
and value (eg: ASPSESSIONXXXXX).
<length> is the maximum number of characters to report in the logs, which
include the cookie name, the equal sign and the value, all in the
standard "name=value" form. The string will be truncated on the
right if it exceeds <length>.
Only the first cookie is captured. Both the "cookie" request headers and the
"set-cookie" response headers are monitored. This is particularly useful to
check for application bugs causing session crossing or stealing between
users, because generally the user's cookies can only change on a login page.
When the cookie was not presented by the client, the associated log column
will report "-". When a request does not cause a cookie to be assigned by the
server, a "-" is reported in the response column.
The capture is performed in the frontend only because it is necessary that
the log format does not change for a given frontend depending on the
backends. This may change in the future. Note that there can be only one
"capture cookie" statement in a frontend. The maximum capture length is set
by the global "tune.http.cookielen" setting and defaults to 63 characters. It
is not possible to specify a capture in a "defaults" section.
Example:
capture cookie ASPSESSION len 32
See also : "capture request header", "capture response header" as well as
section 8 about logging.
capture request header <name> len <length>
Capture and log the last occurrence of the specified request header.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | no
Arguments :
<name> is the name of the header to capture. The header names are not
case-sensitive, but it is a common practice to write them as they
appear in the requests, with the first letter of each word in
upper case. The header name will not appear in the logs, only the
value is reported, but the position in the logs is respected.
<length> is the maximum number of characters to extract from the value and
report in the logs. The string will be truncated on the right if
it exceeds <length>.
The complete value of the last occurrence of the header is captured. The
value will be added to the logs between braces ('{}'). If multiple headers
are captured, they will be delimited by a vertical bar ('|') and will appear
in the same order they were declared in the configuration. Non-existent
headers will be logged just as an empty string. Common uses for request
header captures include the "Host" field in virtual hosting environments, the
"Content-length" when uploads are supported, "User-agent" to quickly
differentiate between real users and robots, and "X-Forwarded-For" in proxied
environments to find where the request came from.
Note that when capturing headers such as "User-agent", some spaces may be
logged, making the log analysis more difficult. Thus be careful about what
you log if you know your log parser is not smart enough to rely on the
braces.
There is no limit to the number of captured request headers nor to their
length, though it is wise to keep them low to limit memory usage per session.
In order to keep log format consistent for a same frontend, header captures
can only be declared in a frontend. It is not possible to specify a capture
in a "defaults" section.
Example:
capture request header Host len 15
capture request header X-Forwarded-For len 15
capture request header Referrer len 15
See also : "capture cookie", "capture response header" as well as section 8
about logging.
capture response header <name> len <length>
Capture and log the last occurrence of the specified response header.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | no
Arguments :
<name> is the name of the header to capture. The header names are not
case-sensitive, but it is a common practice to write them as they
appear in the response, with the first letter of each word in
upper case. The header name will not appear in the logs, only the
value is reported, but the position in the logs is respected.
<length> is the maximum number of characters to extract from the value and
report in the logs. The string will be truncated on the right if
it exceeds <length>.
The complete value of the last occurrence of the header is captured. The
result will be added to the logs between braces ('{}') after the captured
request headers. If multiple headers are captured, they will be delimited by
a vertical bar ('|') and will appear in the same order they were declared in
the configuration. Non-existent headers will be logged just as an empty
string. Common uses for response header captures include the "Content-length"
header which indicates how many bytes are expected to be returned, the
"Location" header to track redirections.
There is no limit to the number of captured response headers nor to their
length, though it is wise to keep them low to limit memory usage per session.
In order to keep log format consistent for a same frontend, header captures
can only be declared in a frontend. It is not possible to specify a capture
in a "defaults" section.
Example:
capture response header Content-length len 9
capture response header Location len 15
See also : "capture cookie", "capture request header" as well as section 8
about logging.
clitimeout <timeout> (deprecated)
Set the maximum inactivity time on the client side.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<timeout> is the timeout value is specified in milliseconds by default, but
can be in any other unit if the number is suffixed by the unit,
as explained at the top of this document.
The inactivity timeout applies when the client is expected to acknowledge or
send data. In HTTP mode, this timeout is particularly important to consider
during the first phase, when the client sends the request, and during the
response while it is reading data sent by the server. The value is specified
in milliseconds by default, but can be in any other unit if the number is
suffixed by the unit, as specified at the top of this document. In TCP mode
(and to a lesser extent, in HTTP mode), it is highly recommended that the
client timeout remains equal to the server timeout in order to avoid complex
situations to debug. It is a good practice to cover one or several TCP packet
losses by specifying timeouts that are slightly above multiples of 3 seconds
(eg: 4 or 5 seconds).
This parameter is specific to frontends, but can be specified once for all in
"defaults" sections. This is in fact one of the easiest solutions not to
forget about it. An unspecified timeout results in an infinite timeout, which
is not recommended. Such a usage is accepted and works but reports a warning
during startup because it may results in accumulation of expired sessions in
the system if the system's timeouts are not configured either.
This parameter is provided for compatibility but is currently deprecated.
Please use "timeout client" instead.
See also : "timeout client", "timeout http-request", "timeout server", and
"srvtimeout".
compression algo <algorithm> ...
compression type <mime type> ...
compression offload
Enable HTTP compression.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
algo is followed by the list of supported compression algorithms.
type is followed by the list of MIME types that will be compressed.
offload makes haproxy work as a compression offloader only (see notes).
The currently supported algorithms are :
identity this is mostly for debugging, and it was useful for developing
the compression feature. Identity does not apply any change on
data.
gzip applies gzip compression. This setting is only available when
support for zlib was built in.
deflate same as gzip, but with deflate algorithm and zlib format.
Note that this algorithm has ambiguous support on many browsers
and no support at all from recent ones. It is strongly
recommended not to use it for anything else than experimentation.
This setting is only available when support for zlib was built
in.
Compression will be activated depending on the Accept-Encoding request
header. With identity, it does not take care of that header.
If backend servers support HTTP compression, these directives
will be no-op: haproxy will see the compressed response and will not
compress again. If backend servers do not support HTTP compression and
there is Accept-Encoding header in request, haproxy will compress the
matching response.
The "offload" setting makes haproxy remove the Accept-Encoding header to
prevent backend servers from compressing responses. It is strongly
recommended not to do this because this means that all the compression work
will be done on the single point where haproxy is located. However in some
deployment scenarios, haproxy may be installed in front of a buggy gateway
with broken HTTP compression implementation which can't be turned off.
In that case haproxy can be used to prevent that gateway from emitting
invalid payloads. In this case, simply removing the header in the
configuration does not work because it applies before the header is parsed,
so that prevents haproxy from compressing. The "offload" setting should
then be used for such scenarios.
Compression is disabled when:
* the request does not advertise a supported compression algorithm in the
"Accept-Encoding" header
* the response message is not HTTP/1.1
* HTTP status code is not 200
* response header "Transfer-Encoding" contains "chunked" (Temporary
Workaround)
* response contain neither a "Content-Length" header nor a
"Transfer-Encoding" whose last value is "chunked"
* response contains a "Content-Type" header whose first value starts with
"multipart"
* the response contains the "no-transform" value in the "Cache-control"
header
* User-Agent matches "Mozilla/4" unless it is MSIE 6 with XP SP2, or MSIE 7
and later
* The response contains a "Content-Encoding" header, indicating that the
response is already compressed (see compression offload)
Note: The compression does not rewrite Etag headers, and does not emit the
Warning header.
Examples :
compression algo gzip
compression type text/html text/plain
contimeout <timeout> (deprecated)
Set the maximum time to wait for a connection attempt to a server to succeed.
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<timeout> is the timeout value is specified in milliseconds by default, but
can be in any other unit if the number is suffixed by the unit,
as explained at the top of this document.
If the server is located on the same LAN as haproxy, the connection should be
immediate (less than a few milliseconds). Anyway, it is a good practice to
cover one or several TCP packet losses by specifying timeouts that are
slightly above multiples of 3 seconds (eg: 4 or 5 seconds). By default, the
connect timeout also presets the queue timeout to the same value if this one
has not been specified. Historically, the contimeout was also used to set the
tarpit timeout in a listen section, which is not possible in a pure frontend.
This parameter is specific to backends, but can be specified once for all in
"defaults" sections. This is in fact one of the easiest solutions not to
forget about it. An unspecified timeout results in an infinite timeout, which
is not recommended. Such a usage is accepted and works but reports a warning
during startup because it may results in accumulation of failed sessions in
the system if the system's timeouts are not configured either.
This parameter is provided for backwards compatibility but is currently
deprecated. Please use "timeout connect", "timeout queue" or "timeout tarpit"
instead.
See also : "timeout connect", "timeout queue", "timeout tarpit",
"timeout server", "contimeout".
cookie <name> [ rewrite | insert | prefix ] [ indirect ] [ nocache ]
[ postonly ] [ preserve ] [ httponly ] [ secure ]
[ domain <domain> ]* [ maxidle <idle> ] [ maxlife <life> ]
Enable cookie-based persistence in a backend.
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<name> is the name of the cookie which will be monitored, modified or
inserted in order to bring persistence. This cookie is sent to
the client via a "Set-Cookie" header in the response, and is
brought back by the client in a "Cookie" header in all requests.
Special care should be taken to choose a name which does not
conflict with any likely application cookie. Also, if the same
backends are subject to be used by the same clients (eg:
HTTP/HTTPS), care should be taken to use different cookie names
between all backends if persistence between them is not desired.
rewrite This keyword indicates that the cookie will be provided by the
server and that haproxy will have to modify its value to set the
server's identifier in it. This mode is handy when the management
of complex combinations of "Set-cookie" and "Cache-control"
headers is left to the application. The application can then
decide whether or not it is appropriate to emit a persistence
cookie. Since all responses should be monitored, this mode only
works in HTTP close mode. Unless the application behaviour is
very complex and/or broken, it is advised not to start with this
mode for new deployments. This keyword is incompatible with
"insert" and "prefix".
insert This keyword indicates that the persistence cookie will have to
be inserted by haproxy in server responses if the client did not
already have a cookie that would have permitted it to access this
server. When used without the "preserve" option, if the server
emits a cookie with the same name, it will be remove before
processing. For this reason, this mode can be used to upgrade
existing configurations running in the "rewrite" mode. The cookie
will only be a session cookie and will not be stored on the
client's disk. By default, unless the "indirect" option is added,
the server will see the cookies emitted by the client. Due to
caching effects, it is generally wise to add the "nocache" or
"postonly" keywords (see below). The "insert" keyword is not
compatible with "rewrite" and "prefix".
prefix This keyword indicates that instead of relying on a dedicated
cookie for the persistence, an existing one will be completed.
This may be needed in some specific environments where the client
does not support more than one single cookie and the application
already needs it. In this case, whenever the server sets a cookie
named <name>, it will be prefixed with the server's identifier
and a delimiter. The prefix will be removed from all client
requests so that the server still finds the cookie it emitted.
Since all requests and responses are subject to being modified,
this mode requires the HTTP close mode. The "prefix" keyword is
not compatible with "rewrite" and "insert". Note: it is highly
recommended not to use "indirect" with "prefix", otherwise server
cookie updates would not be sent to clients.
indirect When this option is specified, no cookie will be emitted to a
client which already has a valid one for the server which has
processed the request. If the server sets such a cookie itself,
it will be removed, unless the "preserve" option is also set. In
"insert" mode, this will additionally remove cookies from the
requests transmitted to the server, making the persistence
mechanism totally transparent from an application point of view.
Note: it is highly recommended not to use "indirect" with
"prefix", otherwise server cookie updates would not be sent to
clients.
nocache This option is recommended in conjunction with the insert mode
when there is a cache between the client and HAProxy, as it
ensures that a cacheable response will be tagged non-cacheable if
a cookie needs to be inserted. This is important because if all
persistence cookies are added on a cacheable home page for
instance, then all customers will then fetch the page from an
outer cache and will all share the same persistence cookie,
leading to one server receiving much more traffic than others.
See also the "insert" and "postonly" options.
postonly This option ensures that cookie insertion will only be performed
on responses to POST requests. It is an alternative to the
"nocache" option, because POST responses are not cacheable, so
this ensures that the persistence cookie will never get cached.
Since most sites do not need any sort of persistence before the
first POST which generally is a login request, this is a very
efficient method to optimize caching without risking to find a
persistence cookie in the cache.
See also the "insert" and "nocache" options.
preserve This option may only be used with "insert" and/or "indirect". It
allows the server to emit the persistence cookie itself. In this
case, if a cookie is found in the response, haproxy will leave it
untouched. This is useful in order to end persistence after a
logout request for instance. For this, the server just has to
emit a cookie with an invalid value (eg: empty) or with a date in
the past. By combining this mechanism with the "disable-on-404"
check option, it is possible to perform a completely graceful
shutdown because users will definitely leave the server after
they logout.
httponly This option tells haproxy to add an "HttpOnly" cookie attribute
when a cookie is inserted. This attribute is used so that a
user agent doesn't share the cookie with non-HTTP components.
Please check RFC6265 for more information on this attribute.
secure This option tells haproxy to add a "Secure" cookie attribute when
a cookie is inserted. This attribute is used so that a user agent
never emits this cookie over non-secure channels, which means
that a cookie learned with this flag will be presented only over
SSL/TLS connections. Please check RFC6265 for more information on
this attribute.
domain This option allows to specify the domain at which a cookie is
inserted. It requires exactly one parameter: a valid domain
name. If the domain begins with a dot, the browser is allowed to
use it for any host ending with that name. It is also possible to
specify several domain names by invoking this option multiple
times. Some browsers might have small limits on the number of
domains, so be careful when doing that. For the record, sending
10 domains to MSIE 6 or Firefox 2 works as expected.
maxidle This option allows inserted cookies to be ignored after some idle
time. It only works with insert-mode cookies. When a cookie is
sent to the client, the date this cookie was emitted is sent too.
Upon further presentations of this cookie, if the date is older
than the delay indicated by the parameter (in seconds), it will
be ignored. Otherwise, it will be refreshed if needed when the
response is sent to the client. This is particularly useful to
prevent users who never close their browsers from remaining for
too long on the same server (eg: after a farm size change). When
this option is set and a cookie has no date, it is always
accepted, but gets refreshed in the response. This maintains the
ability for admins to access their sites. Cookies that have a
date in the future further than 24 hours are ignored. Doing so
lets admins fix timezone issues without risking kicking users off
the site.
maxlife This option allows inserted cookies to be ignored after some life
time, whether they're in use or not. It only works with insert
mode cookies. When a cookie is first sent to the client, the date
this cookie was emitted is sent too. Upon further presentations
of this cookie, if the date is older than the delay indicated by
the parameter (in seconds), it will be ignored. If the cookie in
the request has no date, it is accepted and a date will be set.
Cookies that have a date in the future further than 24 hours are
ignored. Doing so lets admins fix timezone issues without risking
kicking users off the site. Contrary to maxidle, this value is
not refreshed, only the first visit date counts. Both maxidle and
maxlife may be used at the time. This is particularly useful to
prevent users who never close their browsers from remaining for
too long on the same server (eg: after a farm size change). This
is stronger than the maxidle method in that it forces a
redispatch after some absolute delay.
There can be only one persistence cookie per HTTP backend, and it can be
declared in a defaults section. The value of the cookie will be the value
indicated after the "cookie" keyword in a "server" statement. If no cookie
is declared for a given server, the cookie is not set.
Examples :
cookie JSESSIONID prefix
cookie SRV insert indirect nocache
cookie SRV insert postonly indirect
cookie SRV insert indirect nocache maxidle 30m maxlife 8h
See also : "appsession", "balance source", "capture cookie", "server"
and "ignore-persist".
default-server [param*]
Change default options for a server in a backend
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments:
<param*> is a list of parameters for this server. The "default-server"
keyword accepts an important number of options and has a complete
section dedicated to it. Please refer to section 5 for more
details.
Example :
default-server inter 1000 weight 13
See also: "server" and section 5 about server options
default_backend <backend>
Specify the backend to use when no "use_backend" rule has been matched.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<backend> is the name of the backend to use.
When doing content-switching between frontend and backends using the
"use_backend" keyword, it is often useful to indicate which backend will be
used when no rule has matched. It generally is the dynamic backend which
will catch all undetermined requests.
Example :
use_backend dynamic if url_dyn
use_backend static if url_css url_img extension_img
default_backend dynamic
See also : "use_backend", "reqsetbe", "reqisetbe"
description <string>
Describe a listen, frontend or backend.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | yes
Arguments : string
Allows to add a sentence to describe the related object in the HAProxy HTML
stats page. The description will be printed on the right of the object name
it describes.
No need to backslash spaces in the <string> arguments.
disabled
Disable a proxy, frontend or backend.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
The "disabled" keyword is used to disable an instance, mainly in order to
liberate a listening port or to temporarily disable a service. The instance
will still be created and its configuration will be checked, but it will be
created in the "stopped" state and will appear as such in the statistics. It
will not receive any traffic nor will it send any health-checks or logs. It
is possible to disable many instances at once by adding the "disabled"
keyword in a "defaults" section.
See also : "enabled"
dispatch <address>:<port>
Set a default server address
May be used in sections : defaults | frontend | listen | backend
no | no | yes | yes
Arguments :
<address> is the IPv4 address of the default server. Alternatively, a
resolvable hostname is supported, but this name will be resolved
during start-up.
<ports> is a mandatory port specification. All connections will be sent
to this port, and it is not permitted to use port offsets as is
possible with normal servers.
The "dispatch" keyword designates a default server for use when no other
server can take the connection. In the past it was used to forward non
persistent connections to an auxiliary load balancer. Due to its simple
syntax, it has also been used for simple TCP relays. It is recommended not to
use it for more clarity, and to use the "server" directive instead.
See also : "server"
enabled
Enable a proxy, frontend or backend.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
The "enabled" keyword is used to explicitly enable an instance, when the
defaults has been set to "disabled". This is very rarely used.
See also : "disabled"
errorfile <code> <file>
Return a file contents instead of errors generated by HAProxy
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
<code> is the HTTP status code. Currently, HAProxy is capable of
generating codes 200, 400, 403, 408, 500, 502, 503, and 504.
<file> designates a file containing the full HTTP response. It is
recommended to follow the common practice of appending ".http" to
the filename so that people do not confuse the response with HTML
error pages, and to use absolute paths, since files are read
before any chroot is performed.
It is important to understand that this keyword is not meant to rewrite
errors returned by the server, but errors detected and returned by HAProxy.
This is why the list of supported errors is limited to a small set.
Code 200 is emitted in response to requests matching a "monitor-uri" rule.
The files are returned verbatim on the TCP socket. This allows any trick such
as redirections to another URL or site, as well as tricks to clean cookies,
force enable or disable caching, etc... The package provides default error
files returning the same contents as default errors.
The files should not exceed the configured buffer size (BUFSIZE), which
generally is 8 or 16 kB, otherwise they will be truncated. It is also wise
not to put any reference to local contents (eg: images) in order to avoid
loops between the client and HAProxy when all servers are down, causing an
error to be returned instead of an image. For better HTTP compliance, it is
recommended that all header lines end with CR-LF and not LF alone.
The files are read at the same time as the configuration and kept in memory.
For this reason, the errors continue to be returned even when the process is
chrooted, and no file change is considered while the process is running. A
simple method for developing those files consists in associating them to the
403 status code and interrogating a blocked URL.
See also : "errorloc", "errorloc302", "errorloc303"
Example :
errorfile 400 /etc/haproxy/errorfiles/400badreq.http
errorfile 403 /etc/haproxy/errorfiles/403forbid.http
errorfile 503 /etc/haproxy/errorfiles/503sorry.http
errorloc <code> <url>
errorloc302 <code> <url>
Return an HTTP redirection to a URL instead of errors generated by HAProxy
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
<code> is the HTTP status code. Currently, HAProxy is capable of
generating codes 200, 400, 403, 408, 500, 502, 503, and 504.
<url> it is the exact contents of the "Location" header. It may contain
either a relative URI to an error page hosted on the same site,
or an absolute URI designating an error page on another site.
Special care should be given to relative URIs to avoid redirect
loops if the URI itself may generate the same error (eg: 500).
It is important to understand that this keyword is not meant to rewrite
errors returned by the server, but errors detected and returned by HAProxy.
This is why the list of supported errors is limited to a small set.
Code 200 is emitted in response to requests matching a "monitor-uri" rule.
Note that both keyword return the HTTP 302 status code, which tells the
client to fetch the designated URL using the same HTTP method. This can be
quite problematic in case of non-GET methods such as POST, because the URL
sent to the client might not be allowed for something other than GET. To
workaround this problem, please use "errorloc303" which send the HTTP 303
status code, indicating to the client that the URL must be fetched with a GET
request.
See also : "errorfile", "errorloc303"
errorloc303 <code> <url>
Return an HTTP redirection to a URL instead of errors generated by HAProxy
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
<code> is the HTTP status code. Currently, HAProxy is capable of
generating codes 400, 403, 408, 500, 502, 503, and 504.
<url> it is the exact contents of the "Location" header. It may contain
either a relative URI to an error page hosted on the same site,
or an absolute URI designating an error page on another site.
Special care should be given to relative URIs to avoid redirect
loops if the URI itself may generate the same error (eg: 500).
It is important to understand that this keyword is not meant to rewrite
errors returned by the server, but errors detected and returned by HAProxy.
This is why the list of supported errors is limited to a small set.
Code 200 is emitted in response to requests matching a "monitor-uri" rule.
Note that both keyword return the HTTP 303 status code, which tells the
client to fetch the designated URL using the same HTTP GET method. This
solves the usual problems associated with "errorloc" and the 302 code. It is
possible that some very old browsers designed before HTTP/1.1 do not support
it, but no such problem has been reported till now.
See also : "errorfile", "errorloc", "errorloc302"
force-persist { if | unless } <condition>
Declare a condition to force persistence on down servers
May be used in sections: defaults | frontend | listen | backend
no | yes | yes | yes
By default, requests are not dispatched to down servers. It is possible to
force this using "option persist", but it is unconditional and redispatches
to a valid server if "option redispatch" is set. That leaves with very little
possibilities to force some requests to reach a server which is artificially
marked down for maintenance operations.
The "force-persist" statement allows one to declare various ACL-based
conditions which, when met, will cause a request to ignore the down status of
a server and still try to connect to it. That makes it possible to start a
server, still replying an error to the health checks, and run a specially
configured browser to test the service. Among the handy methods, one could
use a specific source IP address, or a specific cookie. The cookie also has
the advantage that it can easily be added/removed on the browser from a test
page. Once the service is validated, it is then possible to open the service
to the world by returning a valid response to health checks.
The forced persistence is enabled when an "if" condition is met, or unless an
"unless" condition is met. The final redispatch is always disabled when this
is used.
See also : "option redispatch", "ignore-persist", "persist",
and section 7 about ACL usage.
fullconn <conns>
Specify at what backend load the servers will reach their maxconn
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<conns> is the number of connections on the backend which will make the
servers use the maximal number of connections.
When a server has a "maxconn" parameter specified, it means that its number
of concurrent connections will never go higher. Additionally, if it has a
"minconn" parameter, it indicates a dynamic limit following the backend's
load. The server will then always accept at least <minconn> connections,
never more than <maxconn>, and the limit will be on the ramp between both
values when the backend has less than <conns> concurrent connections. This
makes it possible to limit the load on the servers during normal loads, but
push it further for important loads without overloading the servers during
exceptional loads.
Since it's hard to get this value right, haproxy automatically sets it to
10% of the sum of the maxconns of all frontends that may branch to this
backend. That way it's safe to leave it unset.
Example :
# The servers will accept between 100 and 1000 concurrent connections each
# and the maximum of 1000 will be reached when the backend reaches 10000
# connections.
backend dynamic
fullconn 10000
server srv1 dyn1:80 minconn 100 maxconn 1000
server srv2 dyn2:80 minconn 100 maxconn 1000
See also : "maxconn", "server"
grace <time>
Maintain a proxy operational for some time after a soft stop
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
<time> is the time (by default in milliseconds) for which the instance
will remain operational with the frontend sockets still listening
when a soft-stop is received via the SIGUSR1 signal.
This may be used to ensure that the services disappear in a certain order.
This was designed so that frontends which are dedicated to monitoring by an
external equipment fail immediately while other ones remain up for the time
needed by the equipment to detect the failure.
Note that currently, there is very little benefit in using this parameter,
and it may in fact complicate the soft-reconfiguration process more than
simplify it.
hash-type <method> <function> <modifier>
Specify a method to use for mapping hashes to servers
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<method> is the method used to select a server from the hash computed by
the <function> :
map-based the hash table is a static array containing all alive servers.
The hashes will be very smooth, will consider weights, but
will be static in that weight changes while a server is up
will be ignored. This means that there will be no slow start.
Also, since a server is selected by its position in the array,
most mappings are changed when the server count changes. This
means that when a server goes up or down, or when a server is
added to a farm, most connections will be redistributed to
different servers. This can be inconvenient with caches for
instance.
consistent the hash table is a tree filled with many occurrences of each
server. The hash key is looked up in the tree and the closest
server is chosen. This hash is dynamic, it supports changing
weights while the servers are up, so it is compatible with the
slow start feature. It has the advantage that when a server
goes up or down, only its associations are moved. When a
server is added to the farm, only a few part of the mappings
are redistributed, making it an ideal method for caches.
However, due to its principle, the distribution will never be
very smooth and it may sometimes be necessary to adjust a
server's weight or its ID to get a more balanced distribution.
In order to get the same distribution on multiple load
balancers, it is important that all servers have the exact
same IDs. Note: consistent hash uses sdbm and avalanche if no
hash function is specified.
<function> is the hash function to be used :
sdbm this function was created intially for sdbm (a public-domain
reimplementation of ndbm) database library. It was found to do
well in scrambling bits, causing better distribution of the keys
and fewer splits. It also happens to be a good general hashing
function with good distribution, unless the total server weight
is a multiple of 64, in which case applying the avalanche
modifier may help.
djb2 this function was first proposed by Dan Bernstein many years ago
on comp.lang.c. Studies have shown that for certain workload this
function provides a better distribution than sdbm. It generally
works well with text-based inputs though it can perform extremely
poorly with numeric-only input or when the total server weight is
a multiple of 33, unless the avalanche modifier is also used.
wt6 this function was designed for haproxy while testing other
functions in the past. It is not as smooth as the other ones, but
is much less sensible to the input data set or to the number of
servers. It can make sense as an alternative to sdbm+avalanche or
djb2+avalanche for consistent hashing or when hashing on numeric
data such as a source IP address or a visitor identifier in a URL
parameter.
<modifier> indicates an optional method applied after hashing the key :
avalanche This directive indicates that the result from the hash
function above should not be used in its raw form but that
a 4-byte full avalanche hash must be applied first. The
purpose of this step is to mix the resulting bits from the
previous hash in order to avoid any undesired effect when
the input contains some limited values or when the number of
servers is a multiple of one of the hash's components (64
for SDBM, 33 for DJB2). Enabling avalanche tends to make the
result less predictable, but it's also not as smooth as when
using the original function. Some testing might be needed
with some workloads. This hash is one of the many proposed
by Bob Jenkins.
The default hash type is "map-based" and is recommended for most usages. The
default function is "sdbm", the selection of a function should be based on
the range of the values being hashed.
See also : "balance", "server"
http-check disable-on-404
Enable a maintenance mode upon HTTP/404 response to health-checks
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
When this option is set, a server which returns an HTTP code 404 will be
excluded from further load-balancing, but will still receive persistent
connections. This provides a very convenient method for Web administrators
to perform a graceful shutdown of their servers. It is also important to note
that a server which is detected as failed while it was in this mode will not
generate an alert, just a notice. If the server responds 2xx or 3xx again, it
will immediately be reinserted into the farm. The status on the stats page
reports "NOLB" for a server in this mode. It is important to note that this
option only works in conjunction with the "httpchk" option. If this option
is used with "http-check expect", then it has precedence over it so that 404
responses will still be considered as soft-stop.
See also : "option httpchk", "http-check expect"
http-check expect [!] <match> <pattern>
Make HTTP health checks consider response contents or specific status codes
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<match> is a keyword indicating how to look for a specific pattern in the
response. The keyword may be one of "status", "rstatus",
"string", or "rstring". The keyword may be preceded by an
exclamation mark ("!") to negate the match. Spaces are allowed
between the exclamation mark and the keyword. See below for more
details on the supported keywords.
<pattern> is the pattern to look for. It may be a string or a regular
expression. If the pattern contains spaces, they must be escaped
with the usual backslash ('\').
By default, "option httpchk" considers that response statuses 2xx and 3xx
are valid, and that others are invalid. When "http-check expect" is used,
it defines what is considered valid or invalid. Only one "http-check"
statement is supported in a backend. If a server fails to respond or times
out, the check obviously fails. The available matches are :
status <string> : test the exact string match for the HTTP status code.
A health check response will be considered valid if the
response's status code is exactly this string. If the
"status" keyword is prefixed with "!", then the response
will be considered invalid if the status code matches.
rstatus <regex> : test a regular expression for the HTTP status code.
A health check response will be considered valid if the
response's status code matches the expression. If the
"rstatus" keyword is prefixed with "!", then the response
will be considered invalid if the status code matches.
This is mostly used to check for multiple codes.
string <string> : test the exact string match in the HTTP response body.
A health check response will be considered valid if the
response's body contains this exact string. If the
"string" keyword is prefixed with "!", then the response
will be considered invalid if the body contains this
string. This can be used to look for a mandatory word at
the end of a dynamic page, or to detect a failure when a
specific error appears on the check page (eg: a stack
trace).
rstring <regex> : test a regular expression on the HTTP response body.
A health check response will be considered valid if the
response's body matches this expression. If the "rstring"
keyword is prefixed with "!", then the response will be
considered invalid if the body matches the expression.
This can be used to look for a mandatory word at the end
of a dynamic page, or to detect a failure when a specific
error appears on the check page (eg: a stack trace).
It is important to note that the responses will be limited to a certain size
defined by the global "tune.chksize" option, which defaults to 16384 bytes.
Thus, too large responses may not contain the mandatory pattern when using
"string" or "rstring". If a large response is absolutely required, it is
possible to change the default max size by setting the global variable.
However, it is worth keeping in mind that parsing very large responses can
waste some CPU cycles, especially when regular expressions are used, and that
it is always better to focus the checks on smaller resources.
Last, if "http-check expect" is combined with "http-check disable-on-404",
then this last one has precedence when the server responds with 404.
Examples :
# only accept status 200 as valid
http-check expect status 200
# consider SQL errors as errors
http-check expect ! string SQL\ Error
# consider status 5xx only as errors
http-check expect ! rstatus ^5
# check that we have a correct hexadecimal tag before /html
http-check expect rstring <!--tag:[0-9a-f]*</html>
See also : "option httpchk", "http-check disable-on-404"
http-check send-state
Enable emission of a state header with HTTP health checks
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
When this option is set, haproxy will systematically send a special header
"X-Haproxy-Server-State" with a list of parameters indicating to each server
how they are seen by haproxy. This can be used for instance when a server is
manipulated without access to haproxy and the operator needs to know whether
haproxy still sees it up or not, or if the server is the last one in a farm.
The header is composed of fields delimited by semi-colons, the first of which
is a word ("UP", "DOWN", "NOLB"), possibly followed by a number of valid
checks on the total number before transition, just as appears in the stats
interface. Next headers are in the form "<variable>=<value>", indicating in
no specific order some values available in the stats interface :
- a variable "name", containing the name of the backend followed by a slash
("/") then the name of the server. This can be used when a server is
checked in multiple backends.
- a variable "node" containing the name of the haproxy node, as set in the
global "node" variable, otherwise the system's hostname if unspecified.
- a variable "weight" indicating the weight of the server, a slash ("/")
and the total weight of the farm (just counting usable servers). This
helps to know if other servers are available to handle the load when this
one fails.
- a variable "scur" indicating the current number of concurrent connections
on the server, followed by a slash ("/") then the total number of
connections on all servers of the same backend.
- a variable "qcur" indicating the current number of requests in the
server's queue.
Example of a header received by the application server :
>>> X-Haproxy-Server-State: UP 2/3; name=bck/srv2; node=lb1; weight=1/2; \
scur=13/22; qcur=0
See also : "option httpchk", "http-check disable-on-404"
http-request { allow | deny | tarpit | auth [realm <realm>] | redirect <rule> |
add-header <name> <fmt> | set-header <name> <fmt> |
set-nice <nice> | set-log-level <level> | set-tos <tos> |
set-mark <mark> }
[ { if | unless } <condition> ]
Access control for Layer 7 requests
May be used in sections: defaults | frontend | listen | backend
no | yes | yes | yes
The http-request statement defines a set of rules which apply to layer 7
processing. The rules are evaluated in their declaration order when they are
met in a frontend, listen or backend section. Any rule may optionally be
followed by an ACL-based condition, in which case it will only be evaluated
if the condition is true.
The first keyword is the rule's action. Currently supported actions include :
- "allow" : this stops the evaluation of the rules and lets the request
pass the check. No further "http-request" rules are evaluated.
- "deny" : this stops the evaluation of the rules and immediately rejects
the request and emits an HTTP 403 error. No further "http-request" rules
are evaluated.
- "tarpit" : this stops the evaluation of the rules and immediately blocks
the request without responding for a delay specified by "timeout tarpit"
or "timeout connect" if the former is not set. After that delay, if the
client is still connected, an HTTP error 500 is returned so that the
client does not suspect it has been tarpitted. Logs will report the flags
"PT". The goal of the tarpit rule is to slow down robots during an attack
when they're limited on the number of concurrent requests. It can be very
efficient against very dumb robots, and will significantly reduce the
load on firewalls compared to a "deny" rule. But when facing "correctly"
developped robots, it can make things worse by forcing haproxy and the
front firewall to support insane number of concurrent connections.
- "auth" : this stops the evaluation of the rules and immediately responds
with an HTTP 401 or 407 error code to invite the user to present a valid
user name and password. No further "http-request" rules are evaluated. An
optional "realm" parameter is supported, it sets the authentication realm
that is returned with the response (typically the application's name).
- "redirect" : this performs an HTTP redirection based on a redirect rule.
This is exactly the same as the "redirect" statement except that it
inserts a redirect rule which can be processed in the middle of other
"http-request" rules and that these rules use the "log-format" strings.
See the "redirect" keyword for the rule's syntax.
- "add-header" appends an HTTP header field whose name is specified in
<name> and whose value is defined by <fmt> which follows the log-format
rules (see Custom Log Format in section 8.2.4). This is particularly
useful to pass connection-specific information to the server (eg: the
client's SSL certificate), or to combine several headers into one. This
rule is not final, so it is possible to add other similar rules. Note
that header addition is performed immediately, so one rule might reuse
the resulting header from a previous rule.
- "set-header" does the same as "add-header" except that the header name
is first removed if it existed. This is useful when passing security
information to the server, where the header must not be manipulated by
external users.
- "set-nice" sets the "nice" factor of the current request being processed.
It only has effect against the other requests being processed at the same
time. The default value is 0, unless altered by the "nice" setting on the
"bind" line. The accepted range is -1024..1024. The higher the value, the
nicest the request will be. Lower values will make the request more
important than other ones. This can be useful to improve the speed of
some requests, or lower the priority of non-important requests. Using
this setting without prior experimentation can cause some major slowdown.
- "set-log-level" is used to change the log level of the current request
when a certain condition is met. Valid levels are the 8 syslog levels
(see the "log" keyword) plus the special level "silent" which disables
logging for this request. This rule is not final so the last matching
rule wins. This rule can be useful to disable health checks coming from
another equipment.
- "set-tos" is used to set the TOS or DSCP field value of packets sent to
the client to the value passed in <tos> on platforms which support this.
This value represents the whole 8 bits of the IP TOS field, and can be
expressed both in decimal or hexadecimal format (prefixed by "0x"). Note
that only the 6 higher bits are used in DSCP or TOS, and the two lower
bits are always 0. This can be used to adjust some routing behaviour on
border routers based on some information from the request. See RFC 2474,
2597, 3260 and 4594 for more information.
- "set-mark" is used to set the Netfilter MARK on all packets sent to the
client to the value passed in <mark> on platforms which support it. This
value is an unsigned 32 bit value which can be matched by netfilter and
by the routing table. It can be expressed both in decimal or hexadecimal
format (prefixed by "0x"). This can be useful to force certain packets to
take a different route (for example a cheaper network path for bulk
downloads). This works on Linux kernels 2.6.32 and above and requires
admin privileges.
There is no limit to the number of http-request statements per instance.
It is important to know that http-request rules are processed very early in
the HTTP processing, just after "block" rules and before "reqdel" or "reqrep"
rules. That way, headers added by "add-header"/"set-header" are visible by
almost all further ACL rules.
Example:
acl nagios src 192.168.129.3
acl local_net src 192.168.0.0/16
acl auth_ok http_auth(L1)
http-request allow if nagios
http-request allow if local_net auth_ok
http-request auth realm Gimme if local_net auth_ok
http-request deny
Example:
acl auth_ok http_auth_group(L1) G1
http-request auth unless auth_ok
Example:
http-request set-header X-Haproxy-Current-Date %T
http-request set-header X-SSL %[ssl_fc]
http-request set-header X-SSL-Session_ID %[ssl_fc_session_id]
http-request set-header X-SSL-Client-Verify %[ssl_c_verify]
http-request set-header X-SSL-Client-DN %{+Q}[ssl_c_s_dn]
http-request set-header X-SSL-Client-CN %{+Q}[ssl_c_s_dn(cn)]
http-request set-header X-SSL-Issuer %{+Q}[ssl_c_i_dn]
http-request set-header X-SSL-Client-NotBefore %{+Q}[ssl_c_notbefore]
http-request set-header X-SSL-Client-NotAfter %{+Q}[ssl_c_notafter]
See also : "stats http-request", section 3.4 about userlists and section 7
about ACL usage.
http-response { allow | deny | add-header <name> <fmt> | set-nice <nice> |
set-header <name> <fmt> | set-log-level <level> |
set-mark <mark> | set-tos <tos> }
[ { if | unless } <condition> ]
Access control for Layer 7 responses
May be used in sections: defaults | frontend | listen | backend
no | yes | yes | yes
The http-response statement defines a set of rules which apply to layer 7
processing. The rules are evaluated in their declaration order when they are
met in a frontend, listen or backend section. Any rule may optionally be
followed by an ACL-based condition, in which case it will only be evaluated
if the condition is true. Since these rules apply on responses, the backend
rules are applied first, followed by the frontend's rules.
The first keyword is the rule's action. Currently supported actions include :
- "allow" : this stops the evaluation of the rules and lets the response
pass the check. No further "http-response" rules are evaluated for the
current section.
- "deny" : this stops the evaluation of the rules and immediately rejects
the response and emits an HTTP 502 error. No further "http-response"
rules are evaluated.
- "add-header" appends an HTTP header field whose name is specified in
<name> and whose value is defined by <fmt> which follows the log-format
rules (see Custom Log Format in section 8.2.4). This may be used to send
a cookie to a client for example, or to pass some internal information.
This rule is not final, so it is possible to add other similar rules.
Note that header addition is performed immediately, so one rule might
reuse the resulting header from a previous rule.
- "set-header" does the same as "add-header" except that the header name
is first removed if it existed. This is useful when passing security
information to the server, where the header must not be manipulated by
external users.
- "set-nice" sets the "nice" factor of the current request being processed.
It only has effect against the other requests being processed at the same
time. The default value is 0, unless altered by the "nice" setting on the
"bind" line. The accepted range is -1024..1024. The higher the value, the
nicest the request will be. Lower values will make the request more
important than other ones. This can be useful to improve the speed of
some requests, or lower the priority of non-important requests. Using
this setting without prior experimentation can cause some major slowdown.
- "set-log-level" is used to change the log level of the current request
when a certain condition is met. Valid levels are the 8 syslog levels
(see the "log" keyword) plus the special level "silent" which disables
logging for this request. This rule is not final so the last matching
rule wins. This rule can be useful to disable health checks coming from
another equipment.
- "set-tos" is used to set the TOS or DSCP field value of packets sent to
the client to the value passed in <tos> on platforms which support this.
This value represents the whole 8 bits of the IP TOS field, and can be
expressed both in decimal or hexadecimal format (prefixed by "0x"). Note
that only the 6 higher bits are used in DSCP or TOS, and the two lower
bits are always 0. This can be used to adjust some routing behaviour on
border routers based on some information from the request. See RFC 2474,
2597, 3260 and 4594 for more information.
- "set-mark" is used to set the Netfilter MARK on all packets sent to the
client to the value passed in <mark> on platforms which support it. This
value is an unsigned 32 bit value which can be matched by netfilter and
by the routing table. It can be expressed both in decimal or hexadecimal
format (prefixed by "0x"). This can be useful to force certain packets to
take a different route (for example a cheaper network path for bulk
downloads). This works on Linux kernels 2.6.32 and above and requires
admin privileges.
There is no limit to the number of http-response statements per instance.
It is important to know that http-response rules are processed very early in
the HTTP processing, before "reqdel" or "reqrep" rules. That way, headers
added by "add-header"/"set-header" are visible by almost all further ACL
rules.
See also : "http-request", section 3.4 about userlists and section 7 about
ACL usage.
tcp-check expect [!] <match> <pattern>
Specify data to be collected and analysed during a generic health check
May be used in sections: defaults | frontend | listen | backend
no | no | yes | yes
Arguments :
<match> is a keyword indicating how to look for a specific pattern in the
response. The keyword may be one of "string", "rstring" or
binary.
The keyword may be preceded by an exclamation mark ("!") to negate
the match. Spaces are allowed between the exclamation mark and the
keyword. See below for more details on the supported keywords.
<pattern> is the pattern to look for. It may be a string or a regular
expression. If the pattern contains spaces, they must be escaped
with the usual backslash ('\').
If the match is set to binary, then the pattern must be passed as
a serie of hexadecimal digits in an even number. Each sequence of
two digits will represent a byte. The hexadecimal digits may be
used upper or lower case.
The available matches are intentionally similar to their http-check cousins :
string <string> : test the exact string matches in the response buffer.
A health check response will be considered valid if the
response's buffer contains this exact string. If the
"string" keyword is prefixed with "!", then the response
will be considered invalid if the body contains this
string. This can be used to look for a mandatory pattern
in a protocol response, or to detect a failure when a
specific error appears in a protocol banner.
rstring <regex> : test a regular expression on the response buffer.
A health check response will be considered valid if the
response's buffer matches this expression. If the
"rstring" keyword is prefixed with "!", then the response
will be considered invalid if the body matches the
expression.
binary <hexstring> : test the exact string in its hexadecimal form matches
in the response buffer. A health check response will
be considered valid if the response's buffer contains
this exact hexadecimal string.
Purpose is to match data on binary protocols.
It is important to note that the responses will be limited to a certain size
defined by the global "tune.chksize" option, which defaults to 16384 bytes.
Thus, too large responses may not contain the mandatory pattern when using
"string", "rstring" or binary. If a large response is absolutely required, it
is possible to change the default max size by setting the global variable.
However, it is worth keeping in mind that parsing very large responses can
waste some CPU cycles, especially when regular expressions are used, and that
it is always better to focus the checks on smaller resources. Also, in its
current state, the check will not find any string nor regex past a null
character in the response. Similarly it is not possible to request matching
the null character.
Examples :
# perform a POP check
option tcp-check
tcp-check expect string +OK\ POP3\ ready
# perform an IMAP check
option tcp-check
tcp-check expect string *\ OK\ IMAP4\ ready
# look for the redis master server
option tcp-check
tcp-check send PING\r\n
tcp-check expect +PONG
tcp-check send info\ replication\r\n
tcp-check expect string role:master
tcp-check send QUIT\r\n
tcp-check expect string +OK
See also : "option tcp-check", "tcp-check send", "http-check expect",
tune.chksize
tcp-check send <data>
Specify a string to be sent as a question during a generic health check
May be used in sections: defaults | frontend | listen | backend
no | no | yes | yes
<data> : the data to be sent as a question during a generic health check
session. For now, <data> must be a string.
Examples :
# look for the redis master server
option tcp-check
tcp-check send info\ replication\r\n
tcp-check expect string role:master
See also : "option tcp-check", "tcp-check expect", "tcp-check send-binary",
tune.chksize
tcp-check send-binary <hexastring>
Specify an hexa digits string to be sent as a binary question during a raw
tcp health check
May be used in sections: defaults | frontend | listen | backend
no | no | yes | yes
<data> : the data to be sent as a question during a generic health check
session. For now, <data> must be a string.
<hexastring> : test the exact string in its hexadecimal form matches in the
response buffer. A health check response will be considered
valid if the response's buffer contains this exact
hexadecimal string.
Purpose is to send binary data to ask on binary protocols.
Examples :
# redis check in binary
option tcp-check
tcp-check send-binary 50494e470d0a # PING\r\n
tcp-check expect binary 2b504F4e47 # +PONG
See also : "option tcp-check", "tcp-check expect", "tcp-check send",
tune.chksize
http-send-name-header [<header>]
Add the server name to a request. Use the header string given by <header>
May be used in sections: defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<header> The header string to use to send the server name
The "http-send-name-header" statement causes the name of the target
server to be added to the headers of an HTTP request. The name
is added with the header string proved.
See also : "server"
id <value>
Set a persistent ID to a proxy.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | yes
Arguments : none
Set a persistent ID for the proxy. This ID must be unique and positive.
An unused ID will automatically be assigned if unset. The first assigned
value will be 1. This ID is currently only returned in statistics.
ignore-persist { if | unless } <condition>
Declare a condition to ignore persistence
May be used in sections: defaults | frontend | listen | backend
no | yes | yes | yes
By default, when cookie persistence is enabled, every requests containing
the cookie are unconditionally persistent (assuming the target server is up
and running).
The "ignore-persist" statement allows one to declare various ACL-based
conditions which, when met, will cause a request to ignore persistence.
This is sometimes useful to load balance requests for static files, which
oftenly don't require persistence. This can also be used to fully disable
persistence for a specific User-Agent (for example, some web crawler bots).
Combined with "appsession", it can also help reduce HAProxy memory usage, as
the appsession table won't grow if persistence is ignored.
The persistence is ignored when an "if" condition is met, or unless an
"unless" condition is met.
See also : "force-persist", "cookie", and section 7 about ACL usage.
log global
log <address> <facility> [<level> [<minlevel>]]
no log
Enable per-instance logging of events and traffic.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Prefix :
no should be used when the logger list must be flushed. For example,
if you don't want to inherit from the default logger list. This
prefix does not allow arguments.
Arguments :
global should be used when the instance's logging parameters are the
same as the global ones. This is the most common usage. "global"
replaces <address>, <facility> and <level> with those of the log
entries found in the "global" section. Only one "log global"
statement may be used per instance, and this form takes no other
parameter.
<address> indicates where to send the logs. It takes the same format as
for the "global" section's logs, and can be one of :
- An IPv4 address optionally followed by a colon (':') and a UDP
port. If no port is specified, 514 is used by default (the
standard syslog port).
- An IPv6 address followed by a colon (':') and optionally a UDP
port. If no port is specified, 514 is used by default (the
standard syslog port).
- A filesystem path to a UNIX domain socket, keeping in mind
considerations for chroot (be sure the path is accessible
inside the chroot) and uid/gid (be sure the path is
appropriately writeable).
Any part of the address string may reference any number of
environment variables by preceding their name with a dollar
sign ('$') and optionally enclosing them with braces ('{}'),
similarly to what is done in Bourne shell.
<facility> must be one of the 24 standard syslog facilities :
kern user mail daemon auth syslog lpr news
uucp cron auth2 ftp ntp audit alert cron2
local0 local1 local2 local3 local4 local5 local6 local7
<level> is optional and can be specified to filter outgoing messages. By
default, all messages are sent. If a level is specified, only
messages with a severity at least as important as this level
will be sent. An optional minimum level can be specified. If it
is set, logs emitted with a more severe level than this one will
be capped to this level. This is used to avoid sending "emerg"
messages on all terminals on some default syslog configurations.
Eight levels are known :
emerg alert crit err warning notice info debug
It is important to keep in mind that it is the frontend which decides what to
log from a connection, and that in case of content switching, the log entries
from the backend will be ignored. Connections are logged at level "info".
However, backend log declaration define how and where servers status changes
will be logged. Level "notice" will be used to indicate a server going up,
"warning" will be used for termination signals and definitive service
termination, and "alert" will be used for when a server goes down.
Note : According to RFC3164, messages are truncated to 1024 bytes before
being emitted.
Example :
log global
log 127.0.0.1:514 local0 notice # only send important events
log 127.0.0.1:514 local0 notice notice # same but limit output level
log ${LOCAL_SYSLOG}:514 local0 notice # send to local server
log-format <string>
Allows you to custom a log line.
See also : Custom Log Format (8.2.4)
maxconn <conns>
Fix the maximum number of concurrent connections on a frontend
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<conns> is the maximum number of concurrent connections the frontend will
accept to serve. Excess connections will be queued by the system
in the socket's listen queue and will be served once a connection
closes.
If the system supports it, it can be useful on big sites to raise this limit
very high so that haproxy manages connection queues, instead of leaving the
clients with unanswered connection attempts. This value should not exceed the
global maxconn. Also, keep in mind that a connection contains two buffers
of 8kB each, as well as some other data resulting in about 17 kB of RAM being
consumed per established connection. That means that a medium system equipped
with 1GB of RAM can withstand around 40000-50000 concurrent connections if
properly tuned.
Also, when <conns> is set to large values, it is possible that the servers
are not sized to accept such loads, and for this reason it is generally wise
to assign them some reasonable connection limits.
By default, this value is set to 2000.
See also : "server", global section's "maxconn", "fullconn"
mode { tcp|http|health }
Set the running mode or protocol of the instance
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
tcp The instance will work in pure TCP mode. A full-duplex connection
will be established between clients and servers, and no layer 7
examination will be performed. This is the default mode. It
should be used for SSL, SSH, SMTP, ...
http The instance will work in HTTP mode. The client request will be
analyzed in depth before connecting to any server. Any request
which is not RFC-compliant will be rejected. Layer 7 filtering,
processing and switching will be possible. This is the mode which
brings HAProxy most of its value.
health The instance will work in "health" mode. It will just reply "OK"
to incoming connections and close the connection. Alternatively,
If the "httpchk" option is set, "HTTP/1.0 200 OK" will be sent
instead. Nothing will be logged in either case. This mode is used
to reply to external components health checks. This mode is
deprecated and should not be used anymore as it is possible to do
the same and even better by combining TCP or HTTP modes with the
"monitor" keyword.
When doing content switching, it is mandatory that the frontend and the
backend are in the same mode (generally HTTP), otherwise the configuration
will be refused.
Example :
defaults http_instances
mode http
See also : "monitor", "monitor-net"
monitor fail { if | unless } <condition>
Add a condition to report a failure to a monitor HTTP request.
May be used in sections : defaults | frontend | listen | backend
no | yes | yes | no
Arguments :
if <cond> the monitor request will fail if the condition is satisfied,
and will succeed otherwise. The condition should describe a
combined test which must induce a failure if all conditions
are met, for instance a low number of servers both in a
backend and its backup.
unless <cond> the monitor request will succeed only if the condition is
satisfied, and will fail otherwise. Such a condition may be
based on a test on the presence of a minimum number of active
servers in a list of backends.
This statement adds a condition which can force the response to a monitor
request to report a failure. By default, when an external component queries
the URI dedicated to monitoring, a 200 response is returned. When one of the
conditions above is met, haproxy will return 503 instead of 200. This is
very useful to report a site failure to an external component which may base
routing advertisements between multiple sites on the availability reported by
haproxy. In this case, one would rely on an ACL involving the "nbsrv"
criterion. Note that "monitor fail" only works in HTTP mode. Both status
messages may be tweaked using "errorfile" or "errorloc" if needed.
Example:
frontend www
mode http
acl site_dead nbsrv(dynamic) lt 2
acl site_dead nbsrv(static) lt 2
monitor-uri /site_alive
monitor fail if site_dead
See also : "monitor-net", "monitor-uri", "errorfile", "errorloc"
monitor-net <source>
Declare a source network which is limited to monitor requests
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<source> is the source IPv4 address or network which will only be able to
get monitor responses to any request. It can be either an IPv4
address, a host name, or an address followed by a slash ('/')
followed by a mask.
In TCP mode, any connection coming from a source matching <source> will cause
the connection to be immediately closed without any log. This allows another
equipment to probe the port and verify that it is still listening, without
forwarding the connection to a remote server.
In HTTP mode, a connection coming from a source matching <source> will be
accepted, the following response will be sent without waiting for a request,
then the connection will be closed : "HTTP/1.0 200 OK". This is normally
enough for any front-end HTTP probe to detect that the service is UP and
running without forwarding the request to a backend server. Note that this
response is sent in raw format, without any transformation. This is important
as it means that it will not be SSL-encrypted on SSL listeners.
Monitor requests are processed very early, just after tcp-request connection
ACLs which are the only ones able to block them. These connections are short
lived and never wait for any data from the client. They cannot be logged, and
it is the intended purpose. They are only used to report HAProxy's health to
an upper component, nothing more. Please note that "monitor fail" rules do
not apply to connections intercepted by "monitor-net".
Last, please note that only one "monitor-net" statement can be specified in
a frontend. If more than one is found, only the last one will be considered.
Example :
# addresses .252 and .253 are just probing us.
frontend www
monitor-net 192.168.0.252/31
See also : "monitor fail", "monitor-uri"
monitor-uri <uri>
Intercept a URI used by external components' monitor requests
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments :
<uri> is the exact URI which we want to intercept to return HAProxy's
health status instead of forwarding the request.
When an HTTP request referencing <uri> will be received on a frontend,
HAProxy will not forward it nor log it, but instead will return either
"HTTP/1.0 200 OK" or "HTTP/1.0 503 Service unavailable", depending on failure
conditions defined with "monitor fail". This is normally enough for any
front-end HTTP probe to detect that the service is UP and running without
forwarding the request to a backend server. Note that the HTTP method, the
version and all headers are ignored, but the request must at least be valid
at the HTTP level. This keyword may only be used with an HTTP-mode frontend.
Monitor requests are processed very early. It is not possible to block nor
divert them using ACLs. They cannot be logged either, and it is the intended
purpose. They are only used to report HAProxy's health to an upper component,
nothing more. However, it is possible to add any number of conditions using
"monitor fail" and ACLs so that the result can be adjusted to whatever check
can be imagined (most often the number of available servers in a backend).
Example :
# Use /haproxy_test to report haproxy's status
frontend www
mode http
monitor-uri /haproxy_test
See also : "monitor fail", "monitor-net"
option abortonclose
no option abortonclose
Enable or disable early dropping of aborted requests pending in queues.
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
In presence of very high loads, the servers will take some time to respond.
The per-instance connection queue will inflate, and the response time will
increase respective to the size of the queue times the average per-session
response time. When clients will wait for more than a few seconds, they will
often hit the "STOP" button on their browser, leaving a useless request in
the queue, and slowing down other users, and the servers as well, because the
request will eventually be served, then aborted at the first error
encountered while delivering the response.
As there is no way to distinguish between a full STOP and a simple output
close on the client side, HTTP agents should be conservative and consider
that the client might only have closed its output channel while waiting for
the response. However, this introduces risks of congestion when lots of users
do the same, and is completely useless nowadays because probably no client at
all will close the session while waiting for the response. Some HTTP agents
support this behaviour (Squid, Apache, HAProxy), and others do not (TUX, most
hardware-based load balancers). So the probability for a closed input channel
to represent a user hitting the "STOP" button is close to 100%, and the risk
of being the single component to break rare but valid traffic is extremely
low, which adds to the temptation to be able to abort a session early while
still not served and not pollute the servers.
In HAProxy, the user can choose the desired behaviour using the option
"abortonclose". By default (without the option) the behaviour is HTTP
compliant and aborted requests will be served. But when the option is
specified, a session with an incoming channel closed will be aborted while
it is still possible, either pending in the queue for a connection slot, or
during the connection establishment if the server has not yet acknowledged
the connection request. This considerably reduces the queue size and the load
on saturated servers when users are tempted to click on STOP, which in turn
reduces the response time for other users.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "timeout queue" and server's "maxconn" and "maxqueue" parameters
option accept-invalid-http-request
no option accept-invalid-http-request
Enable or disable relaxing of HTTP request parsing
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
By default, HAProxy complies with RFC2616 in terms of message parsing. This
means that invalid characters in header names are not permitted and cause an
error to be returned to the client. This is the desired behaviour as such
forbidden characters are essentially used to build attacks exploiting server
weaknesses, and bypass security filtering. Sometimes, a buggy browser or
server will emit invalid header names for whatever reason (configuration,
implementation) and the issue will not be immediately fixed. In such a case,
it is possible to relax HAProxy's header name parser to accept any character
even if that does not make sense, by specifying this option. Similarly, the
list of characters allowed to appear in a URI is well defined by RFC3986, and
chars 0-31, 32 (space), 34 ('"'), 60 ('<'), 62 ('>'), 92 ('\'), 94 ('^'), 96
('`'), 123 ('{'), 124 ('|'), 125 ('}'), 127 (delete) and anything above are
not allowed at all. Haproxy always blocks a number of them (0..32, 127). The
remaining ones are blocked by default unless this option is enabled.
This option should never be enabled by default as it hides application bugs
and open security breaches. It should only be deployed after a problem has
been confirmed.
When this option is enabled, erroneous header names will still be accepted in
requests, but the complete request will be captured in order to permit later
analysis using the "show errors" request on the UNIX stats socket. Similarly,
requests containing invalid chars in the URI part will be logged. Doing this
also helps confirming that the issue has been solved.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option accept-invalid-http-response" and "show errors" on the
stats socket.
option accept-invalid-http-response
no option accept-invalid-http-response
Enable or disable relaxing of HTTP response parsing
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
By default, HAProxy complies with RFC2616 in terms of message parsing. This
means that invalid characters in header names are not permitted and cause an
error to be returned to the client. This is the desired behaviour as such
forbidden characters are essentially used to build attacks exploiting server
weaknesses, and bypass security filtering. Sometimes, a buggy browser or
server will emit invalid header names for whatever reason (configuration,
implementation) and the issue will not be immediately fixed. In such a case,
it is possible to relax HAProxy's header name parser to accept any character
even if that does not make sense, by specifying this option.
This option should never be enabled by default as it hides application bugs
and open security breaches. It should only be deployed after a problem has
been confirmed.
When this option is enabled, erroneous header names will still be accepted in
responses, but the complete response will be captured in order to permit
later analysis using the "show errors" request on the UNIX stats socket.
Doing this also helps confirming that the issue has been solved.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option accept-invalid-http-request" and "show errors" on the
stats socket.
option allbackups
no option allbackups
Use either all backup servers at a time or only the first one
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
By default, the first operational backup server gets all traffic when normal
servers are all down. Sometimes, it may be preferred to use multiple backups
at once, because one will not be enough. When "option allbackups" is enabled,
the load balancing will be performed among all backup servers when all normal
ones are unavailable. The same load balancing algorithm will be used and the
servers' weights will be respected. Thus, there will not be any priority
order between the backup servers anymore.
This option is mostly used with static server farms dedicated to return a
"sorry" page when an application is completely offline.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
option checkcache
no option checkcache
Analyze all server responses and block responses with cacheable cookies
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments : none
Some high-level frameworks set application cookies everywhere and do not
always let enough control to the developer to manage how the responses should
be cached. When a session cookie is returned on a cacheable object, there is a
high risk of session crossing or stealing between users traversing the same
caches. In some situations, it is better to block the response than to let
some sensitive session information go in the wild.
The option "checkcache" enables deep inspection of all server responses for
strict compliance with HTTP specification in terms of cacheability. It
carefully checks "Cache-control", "Pragma" and "Set-cookie" headers in server
response to check if there's a risk of caching a cookie on a client-side
proxy. When this option is enabled, the only responses which can be delivered
to the client are :
- all those without "Set-Cookie" header ;
- all those with a return code other than 200, 203, 206, 300, 301, 410,
provided that the server has not set a "Cache-control: public" header ;
- all those that come from a POST request, provided that the server has not
set a 'Cache-Control: public' header ;
- those with a 'Pragma: no-cache' header
- those with a 'Cache-control: private' header
- those with a 'Cache-control: no-store' header
- those with a 'Cache-control: max-age=0' header
- those with a 'Cache-control: s-maxage=0' header
- those with a 'Cache-control: no-cache' header
- those with a 'Cache-control: no-cache="set-cookie"' header
- those with a 'Cache-control: no-cache="set-cookie,' header
(allowing other fields after set-cookie)
If a response doesn't respect these requirements, then it will be blocked
just as if it was from an "rspdeny" filter, with an "HTTP 502 bad gateway".
The session state shows "PH--" meaning that the proxy blocked the response
during headers processing. Additionally, an alert will be sent in the logs so
that admins are informed that there's something to be fixed.
Due to the high impact on the application, the application should be tested
in depth with the option enabled before going to production. It is also a
good practice to always activate it during tests, even if it is not used in
production, as it will report potentially dangerous application behaviours.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
option clitcpka
no option clitcpka
Enable or disable the sending of TCP keepalive packets on the client side
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
When there is a firewall or any session-aware component between a client and
a server, and when the protocol involves very long sessions with long idle
periods (eg: remote desktops), there is a risk that one of the intermediate
components decides to expire a session which has remained idle for too long.
Enabling socket-level TCP keep-alives makes the system regularly send packets
to the other end of the connection, leaving it active. The delay between
keep-alive probes is controlled by the system only and depends both on the
operating system and its tuning parameters.
It is important to understand that keep-alive packets are neither emitted nor
received at the application level. It is only the network stacks which sees
them. For this reason, even if one side of the proxy already uses keep-alives
to maintain its connection alive, those keep-alive packets will not be
forwarded to the other side of the proxy.
Please note that this has nothing to do with HTTP keep-alive.
Using option "clitcpka" enables the emission of TCP keep-alive probes on the
client side of a connection, which should help when session expirations are
noticed between HAProxy and a client.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option srvtcpka", "option tcpka"
option contstats
Enable continuous traffic statistics updates
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
By default, counters used for statistics calculation are incremented
only when a session finishes. It works quite well when serving small
objects, but with big ones (for example large images or archives) or
with A/V streaming, a graph generated from haproxy counters looks like
a hedgehog. With this option enabled counters get incremented continuously,
during a whole session. Recounting touches a hotpath directly so
it is not enabled by default, as it has small performance impact (~0.5%).
option dontlog-normal
no option dontlog-normal
Enable or disable logging of normal, successful connections
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
There are large sites dealing with several thousand connections per second
and for which logging is a major pain. Some of them are even forced to turn
logs off and cannot debug production issues. Setting this option ensures that
normal connections, those which experience no error, no timeout, no retry nor
redispatch, will not be logged. This leaves disk space for anomalies. In HTTP
mode, the response status code is checked and return codes 5xx will still be
logged.
It is strongly discouraged to use this option as most of the time, the key to
complex issues is in the normal logs which will not be logged here. If you
need to separate logs, see the "log-separate-errors" option instead.
See also : "log", "dontlognull", "log-separate-errors" and section 8 about
logging.
option dontlognull
no option dontlognull
Enable or disable logging of null connections
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
In certain environments, there are components which will regularly connect to
various systems to ensure that they are still alive. It can be the case from
another load balancer as well as from monitoring systems. By default, even a
simple port probe or scan will produce a log. If those connections pollute
the logs too much, it is possible to enable option "dontlognull" to indicate
that a connection on which no data has been transferred will not be logged,
which typically corresponds to those probes.
It is generally recommended not to use this option in uncontrolled
environments (eg: internet), otherwise scans and other malicious activities
would not be logged.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "log", "monitor-net", "monitor-uri" and section 8 about logging.
option forceclose
no option forceclose
Enable or disable active connection closing after response is transferred.
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
Some HTTP servers do not necessarily close the connections when they receive
the "Connection: close" set by "option httpclose", and if the client does not
close either, then the connection remains open till the timeout expires. This
causes high number of simultaneous connections on the servers and shows high
global session times in the logs.
When this happens, it is possible to use "option forceclose". It will
actively close the outgoing server channel as soon as the server has finished
to respond. This option implicitly enables the "httpclose" option. Note that
this option also enables the parsing of the full request and response, which
means we can close the connection to the server very quickly, releasing some
resources earlier than with httpclose.
This option may also be combined with "option http-pretend-keepalive", which
will disable sending of the "Connection: close" header, but will still cause
the connection to be closed once the whole response is received.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option httpclose" and "option http-pretend-keepalive"
option forwardfor [ except <network> ] [ header <name> ] [ if-none ]
Enable insertion of the X-Forwarded-For header to requests sent to servers
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments :
<network> is an optional argument used to disable this option for sources
matching <network>
<name> an optional argument to specify a different "X-Forwarded-For"
header name.
Since HAProxy works in reverse-proxy mode, the servers see its IP address as
their client address. This is sometimes annoying when the client's IP address
is expected in server logs. To solve this problem, the well-known HTTP header
"X-Forwarded-For" may be added by HAProxy to all requests sent to the server.
This header contains a value representing the client's IP address. Since this
header is always appended at the end of the existing header list, the server
must be configured to always use the last occurrence of this header only. See
the server's manual to find how to enable use of this standard header. Note
that only the last occurrence of the header must be used, since it is really
possible that the client has already brought one.
The keyword "header" may be used to supply a different header name to replace
the default "X-Forwarded-For". This can be useful where you might already
have a "X-Forwarded-For" header from a different application (eg: stunnel),
and you need preserve it. Also if your backend server doesn't use the
"X-Forwarded-For" header and requires different one (eg: Zeus Web Servers
require "X-Cluster-Client-IP").
Sometimes, a same HAProxy instance may be shared between a direct client
access and a reverse-proxy access (for instance when an SSL reverse-proxy is
used to decrypt HTTPS traffic). It is possible to disable the addition of the
header for a known source address or network by adding the "except" keyword
followed by the network address. In this case, any source IP matching the
network will not cause an addition of this header. Most common uses are with
private networks or 127.0.0.1.
Alternatively, the keyword "if-none" states that the header will only be
added if it is not present. This should only be used in perfectly trusted
environment, as this might cause a security issue if headers reaching haproxy
are under the control of the end-user.
This option may be specified either in the frontend or in the backend. If at
least one of them uses it, the header will be added. Note that the backend's
setting of the header subargument takes precedence over the frontend's if
both are defined. In the case of the "if-none" argument, if at least one of
the frontend or the backend does not specify it, it wants the addition to be
mandatory, so it wins.
It is important to note that by default, HAProxy works in tunnel mode and
only inspects the first request of a connection, meaning that only the first
request will have the header appended, which is certainly not what you want.
In order to fix this, ensure that any of the "httpclose", "forceclose" or
"http-server-close" options is set when using this option.
Examples :
# Public HTTP address also used by stunnel on the same machine
frontend www
mode http
option forwardfor except 127.0.0.1 # stunnel already adds the header
# Those servers want the IP Address in X-Client
backend www
mode http
option forwardfor header X-Client
See also : "option httpclose", "option http-server-close",
"option forceclose"
option http-keep-alive
no option http-keep-alive
Enable or disable HTTP keep-alive from client to server
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
By default, when a client communicates with a server, HAProxy will only
analyze, log, and process the first request of each connection. Setting
"option http-keep-alive" enables HTTP keep-alive mode on the client- and
server- sides. This provides the lowest latency on the client side (slow
network) and the fastest session reuse on the server side at the expense
of maintaining idle connections to the servers. In general, it is possible
with this option to achieve approximately twice the request rate that the
"http-server-close" option achieves on small objects. There are mainly two
situations where this option may be useful :
- when the server is non-HTTP compliant and authenticates the connection
instead of requests (eg: NTLM authentication)
- when the cost of establishing the connection to the server is significant
compared to the cost of retrieving the associated object from the server.
This last case can happen when the server is a fast static server of cache.
In this case, the server will need to be properly tuned to support high enough
connection counts because connections will last until the client sends another
request.
If the client request has to go to another backend or another server due to
content switching or the load balancing algorithm, the idle connection will
immediately be closed and a new one re-opened. Option "prefer-last-server" is
available to try optimize server selection so that if the server currently
attached to an idle connection is usable, it will be used.
In general it is preferred to use "option http-server-close" with application
servers, and some static servers might benefit from "option http-keep-alive".
At the moment, logs will not indicate whether requests came from the same
session or not. The accept date reported in the logs corresponds to the end
of the previous request, and the request time corresponds to the time spent
waiting for a new request. The keep-alive request time is still bound to the
timeout defined by "timeout http-keep-alive" or "timeout http-request" if
not set.
This option may be set both in a frontend and in a backend. It is enabled if
at least one of the frontend or backend holding a connection has it enabled.
It is worth noting that "option forceclose" and "option http-server-close"
have precedence over "option http-keep-alive".
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option forceclose", "option http-server-close",
"option prefer-last-server", "option http-pretend-keepalive",
"option httpclose", and "1.1. The HTTP transaction model".
option http-no-delay
no option http-no-delay
Instruct the system to favor low interactive delays over performance in HTTP
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
In HTTP, each payload is unidirectional and has no notion of interactivity.
Any agent is expected to queue data somewhat for a reasonably low delay.
There are some very rare server-to-server applications that abuse the HTTP
protocol and expect the payload phase to be highly interactive, with many
interleaved data chunks in both directions within a single request. This is
absolutely not supported by the HTTP specification and will not work across
most proxies or servers. When such applications attempt to do this through
haproxy, it works but they will experience high delays due to the network
optimizations which favor performance by instructing the system to wait for
enough data to be available in order to only send full packets. Typical
delays are around 200 ms per round trip. Note that this only happens with
abnormal uses. Normal uses such as CONNECT requests nor WebSockets are not
affected.
When "option http-no-delay" is present in either the frontend or the backend
used by a connection, all such optimizations will be disabled in order to
make the exchanges as fast as possible. Of course this offers no guarantee on
the functionality, as it may break at any other place. But if it works via
HAProxy, it will work as fast as possible. This option should never be used
by default, and should never be used at all unless such a buggy application
is discovered. The impact of using this option is an increase of bandwidth
usage and CPU usage, which may significantly lower performance in high
latency environments.
option http-pretend-keepalive
no option http-pretend-keepalive
Define whether haproxy will announce keepalive to the server or not
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
When running with "option http-server-close" or "option forceclose", haproxy
adds a "Connection: close" header to the request forwarded to the server.
Unfortunately, when some servers see this header, they automatically refrain
from using the chunked encoding for responses of unknown length, while this
is totally unrelated. The immediate effect is that this prevents haproxy from
maintaining the client connection alive. A second effect is that a client or
a cache could receive an incomplete response without being aware of it, and
consider the response complete.
By setting "option http-pretend-keepalive", haproxy will make the server
believe it will keep the connection alive. The server will then not fall back
to the abnormal undesired above. When haproxy gets the whole response, it
will close the connection with the server just as it would do with the
"forceclose" option. That way the client gets a normal response and the
connection is correctly closed on the server side.
It is recommended not to enable this option by default, because most servers
will more efficiently close the connection themselves after the last packet,
and release its buffers slightly earlier. Also, the added packet on the
network could slightly reduce the overall peak performance. However it is
worth noting that when this option is enabled, haproxy will have slightly
less work to do. So if haproxy is the bottleneck on the whole architecture,
enabling this option might save a few CPU cycles.
This option may be set both in a frontend and in a backend. It is enabled if
at least one of the frontend or backend holding a connection has it enabled.
This option may be combined with "option httpclose", which will cause
keepalive to be announced to the server and close to be announced to the
client. This practice is discouraged though.
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option forceclose", "option http-server-close", and
"option http-keep-alive"
option http-server-close
no option http-server-close
Enable or disable HTTP connection closing on the server side
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | yes
Arguments : none
By default, when a client communicates with a server, HAProxy will only
analyze, log, and process the first request of each connection. Setting
"option http-server-close" enables HTTP connection-close mode on the server
side while keeping the ability to support HTTP keep-alive and pipelining on
the client side. This provides the lowest latency on the client side (slow
network) and the fastest session reuse on the server side to save server
resources, similarly to "option forceclose". It also permits non-keepalive
capable servers to be served in keep-alive mode to the clients if they
conform to the requirements of RFC2616. Please note that some servers do not
always conform to those requirements when they see "Connection: close" in the
request. The effect will be that keep-alive will never be used. A workaround
consists in enabling "option http-pretend-keepalive".
At the moment, logs will not indicate whether requests came from the same
session or not. The accept date reported in the logs corresponds to the end
of the previous request, and the request time corresponds to the time spent
waiting for a new request. The keep-alive request time is still bound to the
timeout defined by "timeout http-keep-alive" or "timeout http-request" if
not set.
This option may be set both in a frontend and in a backend. It is enabled if
at least one of the frontend or backend holding a connection has it enabled.
It is worth noting that "option forceclose" has precedence over "option
http-server-close" and that combining "http-server-close" with "httpclose"
basically achieve the same result as "forceclose".
If this option has been enabled in a "defaults" section, it can be disabled
in a specific instance by prepending the "no" keyword before it.
See also : "option forceclose", "option http-pretend-keepalive",
"option httpclose", "option http-keep-alive", and
"1.1. The HTTP transaction model".
option http-use-proxy-header
no option http-use-proxy-header
Make use of non-standard Proxy-Connection header instead of Connection
May be used in sections : defaults | frontend | listen | backend
yes | yes | yes | no
Arguments : none
While RFC2616 explicitly states that HTTP/1.1 agents must use the
Connection header to indicate their wish of persistent or non-persistent
connections, both browsers and proxies ignore this header for proxied
connections and make use of the undocumented, non-standard Proxy-Connection
header instead. The issue begins when trying to put a load balancer between
browsers and such proxies, because there will be a difference between what
haproxy understands and what the client and the proxy agree on.
By setting this option in a frontend, haproxy can automatically switch to use
that non-standard header if it sees proxied requests. A proxied request is
defined here as one where the URI begins with neither a '/' nor a '*'. The
choice of header only affects requests passing through proxies making use of
one of the "httpclose", "forceclose" and "http-server-close" options. Note
that this option can only be specified in a frontend and will affect the
request along its whole life.
Also, when this option is set, a request which requires authentication will
automatically switch to use proxy authentication headers if it is itself a
proxied request. That makes it possible to check or enforce authentication in
front of an existing proxy.
This option should normally never be used, except in front of a proxy.
See also : "option httpclose", "option forceclose" and "option
http-server-close".
option httpchk
option httpchk <uri>
option httpchk <method> <uri>
option httpchk <method> <uri> <version>
Enable HTTP protocol to check on the servers health
May be used in sections : defaults | frontend | listen | backend
yes | no | yes | yes
Arguments :
<method> is the optional HTTP method used with the requests. When not set,
the "OPTIONS" method is used, as it generally requires low server
processing and is easy to filter out from the logs. Any method
may be used, though it is not recommended to invent non-standard
ones.
<uri> is the URI referenced in the HTTP requests. It defaults to " / "
which is accessible by default on almost any server, but may be
changed to any other URI. Query strings are permitted.
<version> is the optional HTTP version string. It defaults to "HTTP/1.0"
but some servers might behave incorrectly in HTTP 1.0, so turning
it to HTTP/1.1 may sometimes help. Note that the Host field is
mandatory in HTTP/1.1, and as a trick, it is possible to pass it
after "\r\n" following the version string.
By default, server health checks only consist in trying to establish a TCP
connection. When "option httpchk" is specified, a complete HTTP request is