blob: b44c2e0119d9522620087fcfd36e0c02fc4c81f0 [file] [log] [blame]
Willy Tarreaub1a34b62010-05-09 22:37:12 +02001 ----------------------
2 HAProxy how-to
3 ----------------------
Willy Tarreau69eda352012-12-24 16:48:14 +01004 version 1.5-dev16
Willy Tarreaub1a34b62010-05-09 22:37:12 +02005 willy tarreau
Willy Tarreau69eda352012-12-24 16:48:14 +01006 2012/12/24
willy tarreau78345332005-12-18 01:33:16 +01007
8
Willy Tarreaub1a34b62010-05-09 22:37:12 +020091) How to build it
10------------------
11
willy tarreau78345332005-12-18 01:33:16 +010012To build haproxy, you will need :
Willy Tarreaub1a34b62010-05-09 22:37:12 +020013 - GNU make. Neither Solaris nor OpenBSD's make work with the GNU Makefile.
Willy Tarreau83b30c12008-05-25 10:32:50 +020014 However, specific Makefiles for BSD and OSX are provided.
Willy Tarreaub1a34b62010-05-09 22:37:12 +020015 - GCC between 2.91 and 4.5.0. Others may work, but not tested.
willy tarreau78345332005-12-18 01:33:16 +010016 - GNU ld
17
18Also, you might want to build with libpcre support, which will provide a very
Willy Tarreaub1a34b62010-05-09 22:37:12 +020019efficient regex implementation and will also fix some badness on Solaris' one.
willy tarreau78345332005-12-18 01:33:16 +010020
21To build haproxy, you have to choose your target OS amongst the following ones
22and assign it to the TARGET variable :
23
Willy Tarreau83b30c12008-05-25 10:32:50 +020024 - linux22 for Linux 2.2
25 - linux24 for Linux 2.4 and above (default)
26 - linux24e for Linux 2.4 with support for a working epoll (> 0.21)
Willy Tarreau83b30c12008-05-25 10:32:50 +020027 - linux26 for Linux 2.6 and above
Willy Tarreaue0c623d2012-06-04 00:42:09 +020028 - linux2628 for Linux 2.6.28 and above (enables splice and tproxy)
Willy Tarreau83b30c12008-05-25 10:32:50 +020029 - solaris for Solaris 8 or 10 (others untested)
Willy Tarreaub1a34b62010-05-09 22:37:12 +020030 - freebsd for FreeBSD 5 to 8.0 (others untested)
Willy Tarreau3b8e9792012-11-22 00:43:09 +010031 - openbsd for OpenBSD 3.1 to 5.2 (others untested)
Willy Tarreau7dec9652012-06-06 16:15:03 +020032 - aix52 for AIX 5.2
Yitzhak Sapir32087312009-06-14 18:27:54 +020033 - cygwin for Cygwin
Willy Tarreau83b30c12008-05-25 10:32:50 +020034 - generic for any other OS.
35 - custom to manually adjust every setting
willy tarreau78345332005-12-18 01:33:16 +010036
37You may also choose your CPU to benefit from some optimizations. This is
38particularly important on UltraSparc machines. For this, you can assign
39one of the following choices to the CPU variable :
40
41 - i686 for intel PentiumPro, Pentium 2 and above, AMD Athlon
42 - i586 for intel Pentium, AMD K6, VIA C3.
43 - ultrasparc : Sun UltraSparc I/II/III/IV processor
Willy Tarreaua5899aa2010-11-28 07:41:00 +010044 - native : use the build machine's specific processor optimizations
willy tarreau78345332005-12-18 01:33:16 +010045 - generic : any other processor or no specific optimization. (default)
46
Willy Tarreau83b30c12008-05-25 10:32:50 +020047Alternatively, you may just set the CPU_CFLAGS value to the optimal GCC options
48for your platform.
49
Willy Tarreauef7341d2009-04-11 19:45:50 +020050You may want to build specific target binaries which do not match your native
51compiler's target. This is particularly true on 64-bit systems when you want
52to build a 32-bit binary. Use the ARCH variable for this purpose. Right now
Willy Tarreaua5899aa2010-11-28 07:41:00 +010053it only knows about a few x86 variants (i386,i486,i586,i686,x86_64), two
54generic ones (32,64) and sets -m32/-m64 as well as -march=<arch> accordingly.
Willy Tarreauef7341d2009-04-11 19:45:50 +020055
willy tarreau78345332005-12-18 01:33:16 +010056If your system supports PCRE (Perl Compatible Regular Expressions), then you
57really should build with libpcre which is between 2 and 10 times faster than
58other libc implementations. Regex are used for header processing (deletion,
59rewriting, allow, deny). The only inconvenient of libpcre is that it is not
60yet widely spread, so if you build for other systems, you might get into
61trouble if they don't have the dynamic library. In this situation, you should
62statically link libpcre into haproxy so that it will not be necessary to
Willy Tarreau83b30c12008-05-25 10:32:50 +020063install it on target systems. Available build options for PCRE are :
willy tarreau78345332005-12-18 01:33:16 +010064
Willy Tarreau83b30c12008-05-25 10:32:50 +020065 - USE_PCRE=1 to use libpcre, in whatever form is available on your system
willy tarreau78345332005-12-18 01:33:16 +010066 (shared or static)
67
Willy Tarreau83b30c12008-05-25 10:32:50 +020068 - USE_STATIC_PCRE=1 to use a static version of libpcre even if the dynamic
69 one is available. This will enhance portability.
70
Willy Tarreau663148c2012-12-12 00:38:22 +010071 - with no option, use your OS libc's standard regex implementation (default).
Willy Tarreau83b30c12008-05-25 10:32:50 +020072 Warning! group references on Solaris seem broken. Use static-pcre whenever
73 possible.
willy tarreau78345332005-12-18 01:33:16 +010074
Willy Tarreau64bc40b2011-03-23 20:00:53 +010075Recent systems can resolve IPv6 host names using getaddrinfo(). This primitive
76is not present in all libcs and does not work in all of them either. Support in
77glibc was broken before 2.3. Some embedded libs may not properly work either,
78thus, support is disabled by default, meaning that some host names which only
79resolve as IPv6 addresses will not resolve and configs might emit an error
80during parsing. If you know that your OS libc has reliable support for
81getaddrinfo(), you can add USE_GETADDRINFO=1 on the make command line to enable
82it. This is the recommended option for most Linux distro packagers since it's
83working fine on all recent mainstream distros. It is automatically enabled on
84Solaris 8 and above, as it's known to work.
85
Willy Tarreaud4508812012-09-10 09:07:41 +020086It is possible to add native support for SSL using the GNU makefile only, and
87by passing "USE_OPENSSL=1" on the make commande line. The libssl and libcrypto
88will automatically be linked with haproxy. Some systems also require libz, so
89if the build fails due to missing symbols such as deflateInit(), then try again
90with "ADDLIB=-lz".
91
William Lallemand82fe75c2012-10-23 10:25:10 +020092It is also possible to include native support for ZLIB to benefit from HTTP
93compression. For this, pass "USE_ZLIB=1" on the "make" command line and ensure
94that zlib is present on the system.
95
willy tarreau78345332005-12-18 01:33:16 +010096By default, the DEBUG variable is set to '-g' to enable debug symbols. It is
97not wise to disable it on uncommon systems, because it's often the only way to
98get a complete core when you need one. Otherwise, you can set DEBUG to '-s' to
99strip the binary.
100
101For example, I use this to build for Solaris 8 :
102
Willy Tarreau83b30c12008-05-25 10:32:50 +0200103 $ make TARGET=solaris CPU=ultrasparc USE_STATIC_PCRE=1
willy tarreau78345332005-12-18 01:33:16 +0100104
Willy Tarreau83b30c12008-05-25 10:32:50 +0200105And I build it this way on OpenBSD or FreeBSD :
willy tarreaud38e72d2006-03-19 20:56:52 +0100106
107 $ make -f Makefile.bsd REGEX=pcre DEBUG= COPTS.generic="-Os -fomit-frame-pointer -mgnu"
108
Willy Tarreau663148c2012-12-12 00:38:22 +0100109And on a classic Linux with SSL and ZLIB support (eg: Red Hat 5.x) :
110
111 $ make TARGET=linux26 CPU=native USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1
112
113And on a recent Linux >= 2.6.28 with SSL and ZLIB support :
Willy Tarreaud4508812012-09-10 09:07:41 +0200114
William Lallemand82fe75c2012-10-23 10:25:10 +0200115 $ make TARGET=linux2628 CPU=native USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1
Willy Tarreaud4508812012-09-10 09:07:41 +0200116
William Lallemand82fe75c2012-10-23 10:25:10 +0200117In order to build a 32-bit binary on an x86_64 Linux system with SSL support
118without support for compression but when OpenSSL requires ZLIB anyway :
Willy Tarreauef7341d2009-04-11 19:45:50 +0200119
Willy Tarreaud4508812012-09-10 09:07:41 +0200120 $ make TARGET=linux26 ARCH=i386 USE_OPENSSL=1 ADDLIB=-lz
Willy Tarreauef7341d2009-04-11 19:45:50 +0200121
Willy Tarreau3b8e9792012-11-22 00:43:09 +0100122The BSD and OSX makefiles do not support build options for OpenSSL nor zlib.
123Also, at least on OpenBSD, pthread_mutexattr_setpshared() does not exist so
124the SSL session cache cannot be shared between multiple processes. If you want
125to enable these options, you need to use GNU make with the default makefile as
126follows :
127
128 $ gmake TARGET=openbsd USE_OPENSSL=1 USE_ZLIB=1 USE_PRIVATE_CACHE=1
129
willy tarreau78345332005-12-18 01:33:16 +0100130If you need to pass other defines, includes, libraries, etc... then please
131check the Makefile to see which ones will be available in your case, and
Willy Tarreau83b30c12008-05-25 10:32:50 +0200132use the USE_* variables in the GNU Makefile, or ADDINC, ADDLIB, and DEFINE
133variables in the BSD makefiles.
willy tarreau78345332005-12-18 01:33:16 +0100134
Willy Tarreau97ec9692010-01-28 20:52:05 +0100135AIX 5.3 is known to work with the generic target. However, for the binary to
136also run on 5.2 or earlier, you need to build with DEFINE="-D_MSGQSUPPORT",
137otherwise __fd_select() will be used while not being present in the libc.
Willy Tarreau7dec9652012-06-06 16:15:03 +0200138If you get build errors because of strange symbols or section mismatches,
139simply remove -g from DEBUG_CFLAGS.
Willy Tarreau97ec9692010-01-28 20:52:05 +0100140
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200141
1422) How to install it
143--------------------
144
145To install haproxy, you can either copy the single resulting binary to the
146place you want, or run :
147
148 $ sudo make install
149
150If you're packaging it for another system, you can specify its root directory
151in the usual DESTDIR variable.
152
153
1543) How to set it up
155-------------------
156
157There is some documentation in the doc/ directory :
158
159 - architecture.txt : this is the architecture manual. It is quite old and
160 does not tell about the nice new features, but it's still a good starting
161 point when you know what you want but don't know how to do it.
162
163 - configuration.txt : this is the configuration manual. It recalls a few
164 essential HTTP basic concepts, and details all the configuration file
165 syntax (keywords, units). It also describes the log and stats format. It
166 is normally always up to date. If you see that something is missing from
167 it, please report it as this is a bug.
168
169 - haproxy-en.txt / haproxy-fr.txt : these are the old outdated docs. You
170 should never need them. If you do, then please report what you didn't
171 find in the other ones.
172
173 - gpl.txt / lgpl.txt : the copy of the licenses covering the software. See
174 the 'LICENSE' file at the top for more information.
175
176 - the rest is mainly for developers.
177
178There are also a number of nice configuration examples in the "examples"
179directory as well as on several sites and articles on the net which are linked
180to from the haproxy web site.
181
182
1834) How to report a bug
184----------------------
185
186It is possible that from time to time you'll find a bug. A bug is a case where
187what you see is not what is documented. Otherwise it can be a misdesign. If you
188find that something is stupidly design, please discuss it on the list (see the
189"how to contribute" section below). If you feel like you're proceeding right
190and haproxy doesn't obey, then first ask yourself if it is possible that nobody
191before you has even encountered this issue. If it's unlikely, the you probably
192have an issue in your setup. Just in case of doubt, please consult the mailing
193list archives :
194
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200195 http://marc.info/?l=haproxy
196
197Otherwise, please try to gather the maximum amount of information to help
198reproduce the issue and send that to the mailing list :
199
200 haproxy@formilux.org
201
202Please include your configuration and logs. You can mask your IP addresses and
203passwords, we don't need them. But it's essential that you post your config if
204you want people to guess what is happening.
205
206Also, keep in mind that haproxy is designed to NEVER CRASH. If you see it die
207without any reason, then it definitely is a critical bug that must be reported
208and urgently fixed. It has happened a couple of times in the past, essentially
209on development versions running on new architectures. If you think your setup
210is fairly common, then it is possible that the issue is totally unrelated.
211Anyway, if that happens, feel free to contact me directly, as I will give you
212instructions on how to collect a usable core file, and will probably ask for
213other captures that you'll not want to share with the list.
214
215
2165) How to contribute
217--------------------
218
219It is possible that you'll want to add a specific feature to satisfy your needs
220or one of your customers'. Contributions are welcome, however I'm often very
221picky about changes. I will generally reject patches that change massive parts
222of the code, or that touch the core parts without any good reason if those
223changes have not been discussed first.
224
225The proper place to discuss your changes is the HAProxy Mailing List. There are
226enough skilled readers to catch hazardous mistakes and to suggest improvements.
Willy Tarreau9a639a12011-09-10 22:48:36 +0200227I trust a number of them enough to merge a patch if they say it's OK, so using
228the list is the fastest way to get your code reviewed and merged. You can
229subscribe to it by sending an empty e-mail at the following address :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200230
231 haproxy+subscribe@formilux.org
232
Willy Tarreau9a639a12011-09-10 22:48:36 +0200233If you have an idea about something to implement, *please* discuss it on the
234list first. It has already happened several times that two persons did the same
235thing simultaneously. This is a waste of time for both of them. It's also very
236common to see some changes rejected because they're done in a way that will
237conflict with future evolutions, or that does not leave a good feeling. It's
238always unpleasant for the person who did the work, and it is unpleasant for me
239too because I value people's time and efforts. That would not happen if these
240were discussed first. There is no problem posting work in progress to the list,
241it happens quite often in fact. Also, don't waste your time with the doc when
242submitting patches for review, only add the doc with the patch you consider
243ready to merge.
244
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200245If your work is very confidential and you can't publicly discuss it, you can
246also mail me directly about it, but your mail may be waiting several days in
247the queue before you get a response.
248
249If you'd like a feature to be added but you think you don't have the skills to
250implement it yourself, you should follow these steps :
251
252 1. discuss the feature on the mailing list. It is possible that someone
253 else has already implemented it, or that someone will tell you how to
254 proceed without it, or even why not to do it. It is also possible that
255 in fact it's quite easy to implement and people will guide you through
256 the process. That way you'll finally have YOUR patch merged, providing
257 the feature YOU need.
258
259 2. if you really can't code it yourself after discussing it, then you may
260 consider contacting someone to do the job for you. Some people on the
261 list might be OK with trying to do it. Otherwise, you can check the list
262 of contributors at the URL below, some of the regular contributors may
263 be able to do the work, probably not for free but their time is as much
264 valuable as yours after all, you can't eat the cake and have it too.
265
266The list of past and regular contributors is available below. It lists not only
267significant code contributions (features, fixes), but also time or money
268donations :
269
270 http://haproxy.1wt.eu/contrib.html
271
272Note to contributors: it's very handy when patches comes with a properly
Willy Tarreau9a639a12011-09-10 22:48:36 +0200273formated subject. There are 3 criteria of particular importance in any patch :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200274
Willy Tarreau9a639a12011-09-10 22:48:36 +0200275 - its nature (is it a fix for a bug, a new feature, an optimization, ...)
276 - its importance, which generally reflects the risk of merging/not merging it
277 - what area it applies to (eg: http, stats, startup, config, doc, ...)
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200278
Willy Tarreau9a639a12011-09-10 22:48:36 +0200279It's important to make these 3 criteria easy to spot in the patch's subject,
280because it's the first (and sometimes the only) thing which is read when
281reviewing patches to find which ones need to be backported to older versions.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200282
Willy Tarreau9a639a12011-09-10 22:48:36 +0200283Specifically, bugs must be clearly easy to spot so that they're never missed.
284Any patch fixing a bug must have the "BUG" tag in its subject. Most common
285patch types include :
286
287 - BUG fix for a bug. The severity of the bug should also be indicated
288 when known. Similarly, if a backport is needed to older versions,
289 it should be indicated on the last line of the commit message. If
290 the bug has been identified as a regression brought by a specific
291 patch or version, this indication will be appreciated too. New
292 maintenance releases are generally emitted when a few of these
293 patches are merged.
294
295 - CLEANUP code cleanup, silence of warnings, etc... theorically no impact.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200296 These patches will rarely be seen in stable branches, though they
Willy Tarreau9a639a12011-09-10 22:48:36 +0200297 may appear when they remove some annoyance or when they make
298 backporting easier. By nature, a cleanup is always minor.
299
300 - REORG code reorganization. Some blocks may be moved to other places,
301 some important checks might be swapped, etc... These changes
302 always present a risk of regression. For this reason, they should
303 never be mixed with any bug fix nor functional change. Code is
304 only moved as-is. Indicating the risk of breakage is highly
305 recommended.
306
307 - BUILD updates or fixes for build issues. Changes to makefiles also fall
308 into this category. The risk of breakage should be indicated if
309 known. It is also appreciated to indicate what platforms and/or
310 configurations were tested after the change.
311
312 - OPTIM some code was optimised. Sometimes if the regression risk is very
313 low and the gains significant, such patches may be merged in the
314 stable branch. Depending on the amount of code changed or replaced
315 and the level of trust the author has in the change, the risk of
316 regression should be indicated.
317
318 - RELEASE release of a new version (development or stable).
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200319
Willy Tarreau9a639a12011-09-10 22:48:36 +0200320 - LICENSE licensing updates (may impact distro packagers).
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200321
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200322
Willy Tarreau9a639a12011-09-10 22:48:36 +0200323When the patch cannot be categorized, it's best not to put any tag. This is
324commonly the case for new features, which development versions are mostly made
325of.
326
327Additionally, the importance of the patch should be indicated when known. A
328single upper-case word is preferred, among :
329
330 - MINOR minor change, very low risk of impact. It is often the case for
331 code additions that don't touch live code. For a bug, it generally
332 indicates an annoyance, nothing more.
333
334 - MEDIUM medium risk, may cause unexpected regressions of low importance or
335 which may quickly be discovered. For a bug, it generally indicates
336 something odd which requires changing the configuration in an
337 undesired way to work around the issue.
338
339 - MAJOR major risk of hidden regression. This happens when I rearrange
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200340 large parts of code, when I play with timeouts, with variable
341 initializations, etc... We should only exceptionally find such
Willy Tarreau9a639a12011-09-10 22:48:36 +0200342 patches in stable branches. For a bug, it indicates severe
343 reliability issues for which workarounds are identified with or
344 without performance impacts.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200345
Willy Tarreau9a639a12011-09-10 22:48:36 +0200346 - CRITICAL medium-term reliability or security is at risk and workarounds,
347 if they exist, might not always be acceptable. An upgrade is
348 absolutely required. A maintenance release may be emitted even if
349 only one of these bugs are fixed. Note that this tag is only used
350 with bugs. Such patches must indicate what is the first version
351 affected, and if known, the commit ID which introduced the issue.
352
353If this criterion doesn't apply, it's best not to put it. For instance, most
354doc updates and most examples or test files are just added or updated without
355any need to qualify a level of importance.
356
357The area the patch applies to is quite important, because some areas are known
358to be similar in older versions, suggesting a backport might be desirable, and
359conversely, some areas are known to be specific to one version. When the tag is
360used alone, uppercase is preferred for readability, otherwise lowercase is fine
361too. The following tags are suggested but not limitative :
362
363 - doc documentation updates or fixes. No code is affected, no need to
364 upgrade. These patches can also be sent right after a new feature,
365 to document it.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200366
Willy Tarreau9a639a12011-09-10 22:48:36 +0200367 - examples example files. Be careful, sometimes these files are packaged.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200368
Willy Tarreau9a639a12011-09-10 22:48:36 +0200369 - tests regression test files. No code is affected, no need to upgrade.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200370
Willy Tarreau9a639a12011-09-10 22:48:36 +0200371 - init initialization code, arguments parsing, etc...
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200372
Willy Tarreau9a639a12011-09-10 22:48:36 +0200373 - config configuration parser, mostly used when adding new config keywords
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200374
Willy Tarreau9a639a12011-09-10 22:48:36 +0200375 - http the HTTP engine
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200376
Willy Tarreau9a639a12011-09-10 22:48:36 +0200377 - stats the stats reporting engine as well as the stats socket CLI
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200378
Willy Tarreau9a639a12011-09-10 22:48:36 +0200379 - checks the health checks engine (eg: when adding new checks)
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200380
Willy Tarreau9a639a12011-09-10 22:48:36 +0200381 - acl the ACL processing core or some ACLs from other areas
382
383 - peers the peer synchronization engine
384
385 - listeners everything related to incoming connection settings
386
387 - frontend everything related to incoming connection processing
388
389 - backend everything related to LB algorithms and server farm
390
391 - session session processing and flags (very sensible, be careful)
392
393 - server server connection management, queueing
394
395 - proxy proxy maintenance (start/stop)
396
397 - log log management
398
399 - poll any of the pollers
400
401 - halog the halog sub-component in the contrib directory
402
403 - contrib any addition to the contrib directory
404
405Other names may be invented when more precise indications are meaningful, for
406instance : "cookie" which indicates cookie processing in the HTTP core. Last,
407indicating the name of the affected file is also a good way to quickly spot
408changes. Many commits were already tagged with "stream_sock" or "cfgparse" for
409instance.
410
411It is desired that AT LEAST one of the 3 criteria tags is reported in the patch
412subject. Ideally, we would have the 3 most often. The two first criteria should
413be present before a first colon (':'). If both are present, then they should be
414delimited with a slash ('/'). The 3rd criterion (area) should appear next, also
415followed by a colon. Thus, all of the following messages are valid :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200416
417Examples of messages :
Willy Tarreau9a639a12011-09-10 22:48:36 +0200418 - DOC: document options forwardfor to logasap
419 - DOC/MAJOR: reorganize the whole document and change indenting
420 - BUG: stats: connection reset counters must be plain ascii, not HTML
421 - BUG/MINOR: stats: connection reset counters must be plain ascii, not HTML
422 - MEDIUM: checks: support multi-packet health check responses
423 - RELEASE: Released version 1.4.2
424 - BUILD: stats: stdint is not present on solaris
425 - OPTIM/MINOR: halog: make fgets parse more bytes by blocks
426 - REORG/MEDIUM: move syscall redefinition to specific places
427
428Please do not use square brackets anymore around the tags, because they give me
429more work when merging patches. By default I'm asking Git to keep them but this
430causes trouble when patches are prefixed with the [PATCH] tag because in order
431not to store it, I have to hand-edit the patches. So as of now, I will ask Git
432to remove whatever is located between square brackets, which implies that any
433subject formatted the old way will have its tag stripped out.
434
435In fact, one of the only square bracket tags that still makes sense is '[RFC]'
436at the beginning of the subject, when you're asking for someone to review your
437change before getting it merged. If the patch is OK to be merged, then I can
438merge it as-is and the '[RFC]' tag will automatically be removed. If you don't
439want it to be merged at all, you can simply state it in the message, or use an
440alternate '[WIP]' tag ("work in progress").
441
442The tags are not rigid, follow your intuition first, anyway I reserve the right
443to change them when merging the patch. It may happen that a same patch has a
444different tag in two distinct branches. The reason is that a bug in one branch
445may just be a cleanup in the other one because the code cannot be triggered.
446
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200447
448For a more efficient interaction between the mainline code and your code, I can
449only strongly encourage you to try the Git version control system :
450
451 http://git-scm.com/
452
453It's very fast, lightweight and lets you undo/redo your work as often as you
454want, without making your mistakes visible to the rest of the world. It will
455definitely help you contribute quality code and take other people's feedback
456in consideration. In order to clone the HAProxy Git repository :
457
458 $ git clone http://git.1wt.eu/git/haproxy-1.4.git (stable 1.4)
459 $ git clone http://git.1wt.eu/git/haproxy.git/ (development)
460
Willy Tarreau663148c2012-12-12 00:38:22 +0100461The site above is slow, a faster mirror is maintained up to date here :
462
463 $ git clone http://master.formilux.org/git/people/willy/haproxy.git/
464
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200465If you decide to use Git for your developments, then your commit messages will
466have the subject line in the format described above, then the whole description
467of your work (mainly why you did it) will be in the body. You can directly send
468your commits to the mailing list, the format is convenient to read and process.
469
willy tarreau78345332005-12-18 01:33:16 +0100470-- end