blob: 4a43da125000f97c1eae970f303875fff294eb74 [file] [log] [blame]
Willy Tarreaub1a34b62010-05-09 22:37:12 +02001 ----------------------
2 HAProxy how-to
3 ----------------------
Willy Tarreau1a34d572014-02-03 00:41:29 +01004 version 1.5-dev22
Willy Tarreaub1a34b62010-05-09 22:37:12 +02005 willy tarreau
Willy Tarreau1a34d572014-02-03 00:41:29 +01006 2014/02/03
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 Tarreau32e65ef2013-04-02 08:14:29 +020015 - GCC between 2.91 and 4.7. 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 Tarreau8624cab2013-04-02 08:17:43 +020031 - osx for Mac OS/X
Willy Tarreau3b8e9792012-11-22 00:43:09 +010032 - openbsd for OpenBSD 3.1 to 5.2 (others untested)
Willy Tarreau50abe302014-04-02 20:44:43 +020033 - aix51 for AIX 5.1
Willy Tarreau7dec9652012-06-06 16:15:03 +020034 - aix52 for AIX 5.2
Yitzhak Sapir32087312009-06-14 18:27:54 +020035 - cygwin for Cygwin
Willy Tarreau83b30c12008-05-25 10:32:50 +020036 - generic for any other OS.
37 - custom to manually adjust every setting
willy tarreau78345332005-12-18 01:33:16 +010038
39You may also choose your CPU to benefit from some optimizations. This is
40particularly important on UltraSparc machines. For this, you can assign
41one of the following choices to the CPU variable :
42
43 - i686 for intel PentiumPro, Pentium 2 and above, AMD Athlon
44 - i586 for intel Pentium, AMD K6, VIA C3.
45 - ultrasparc : Sun UltraSparc I/II/III/IV processor
Willy Tarreaua5899aa2010-11-28 07:41:00 +010046 - native : use the build machine's specific processor optimizations
willy tarreau78345332005-12-18 01:33:16 +010047 - generic : any other processor or no specific optimization. (default)
48
Willy Tarreau83b30c12008-05-25 10:32:50 +020049Alternatively, you may just set the CPU_CFLAGS value to the optimal GCC options
50for your platform.
51
Willy Tarreauef7341d2009-04-11 19:45:50 +020052You may want to build specific target binaries which do not match your native
53compiler's target. This is particularly true on 64-bit systems when you want
54to build a 32-bit binary. Use the ARCH variable for this purpose. Right now
Willy Tarreaua5899aa2010-11-28 07:41:00 +010055it only knows about a few x86 variants (i386,i486,i586,i686,x86_64), two
56generic ones (32,64) and sets -m32/-m64 as well as -march=<arch> accordingly.
Willy Tarreauef7341d2009-04-11 19:45:50 +020057
willy tarreau78345332005-12-18 01:33:16 +010058If your system supports PCRE (Perl Compatible Regular Expressions), then you
59really should build with libpcre which is between 2 and 10 times faster than
60other libc implementations. Regex are used for header processing (deletion,
61rewriting, allow, deny). The only inconvenient of libpcre is that it is not
62yet widely spread, so if you build for other systems, you might get into
63trouble if they don't have the dynamic library. In this situation, you should
64statically link libpcre into haproxy so that it will not be necessary to
Willy Tarreau83b30c12008-05-25 10:32:50 +020065install it on target systems. Available build options for PCRE are :
willy tarreau78345332005-12-18 01:33:16 +010066
Willy Tarreau83b30c12008-05-25 10:32:50 +020067 - USE_PCRE=1 to use libpcre, in whatever form is available on your system
willy tarreau78345332005-12-18 01:33:16 +010068 (shared or static)
69
Willy Tarreau83b30c12008-05-25 10:32:50 +020070 - USE_STATIC_PCRE=1 to use a static version of libpcre even if the dynamic
71 one is available. This will enhance portability.
72
Willy Tarreau663148c2012-12-12 00:38:22 +010073 - with no option, use your OS libc's standard regex implementation (default).
Willy Tarreau83b30c12008-05-25 10:32:50 +020074 Warning! group references on Solaris seem broken. Use static-pcre whenever
75 possible.
willy tarreau78345332005-12-18 01:33:16 +010076
Willy Tarreau64bc40b2011-03-23 20:00:53 +010077Recent systems can resolve IPv6 host names using getaddrinfo(). This primitive
78is not present in all libcs and does not work in all of them either. Support in
79glibc was broken before 2.3. Some embedded libs may not properly work either,
80thus, support is disabled by default, meaning that some host names which only
81resolve as IPv6 addresses will not resolve and configs might emit an error
82during parsing. If you know that your OS libc has reliable support for
83getaddrinfo(), you can add USE_GETADDRINFO=1 on the make command line to enable
84it. This is the recommended option for most Linux distro packagers since it's
85working fine on all recent mainstream distros. It is automatically enabled on
86Solaris 8 and above, as it's known to work.
87
Willy Tarreaud4508812012-09-10 09:07:41 +020088It is possible to add native support for SSL using the GNU makefile only, and
89by passing "USE_OPENSSL=1" on the make commande line. The libssl and libcrypto
90will automatically be linked with haproxy. Some systems also require libz, so
91if the build fails due to missing symbols such as deflateInit(), then try again
92with "ADDLIB=-lz".
93
Lukas Tribus3fe9f1e2013-05-19 16:28:17 +020094To link OpenSSL statically against haproxy, build OpenSSL with the no-shared
95keyword and install it to a local directory, so your system is not affected :
96
97 $ export STATICLIBSSL=/tmp/staticlibssl
98 $ ./config --prefix=$STATICLIBSSL no-shared
99 $ make && make install_sw
100
Lukas Tribus130ddf72013-10-01 00:28:03 +0200101When building haproxy, pass that path via SSL_INC and SSL_LIB to make and
102include additional libs with ADDLIB if needed (in this case for example libdl):
103 $ make TARGET=linux26 USE_OPENSSL=1 SSL_INC=$STATICLIBSSL/include SSL_LIB=$STATICLIBSSL/lib ADDLIB=-ldl
Lukas Tribus3fe9f1e2013-05-19 16:28:17 +0200104
William Lallemand82fe75c2012-10-23 10:25:10 +0200105It is also possible to include native support for ZLIB to benefit from HTTP
106compression. For this, pass "USE_ZLIB=1" on the "make" command line and ensure
107that zlib is present on the system.
108
willy tarreau78345332005-12-18 01:33:16 +0100109By default, the DEBUG variable is set to '-g' to enable debug symbols. It is
110not wise to disable it on uncommon systems, because it's often the only way to
111get a complete core when you need one. Otherwise, you can set DEBUG to '-s' to
112strip the binary.
113
114For example, I use this to build for Solaris 8 :
115
Willy Tarreau83b30c12008-05-25 10:32:50 +0200116 $ make TARGET=solaris CPU=ultrasparc USE_STATIC_PCRE=1
willy tarreau78345332005-12-18 01:33:16 +0100117
Willy Tarreau83b30c12008-05-25 10:32:50 +0200118And I build it this way on OpenBSD or FreeBSD :
willy tarreaud38e72d2006-03-19 20:56:52 +0100119
120 $ make -f Makefile.bsd REGEX=pcre DEBUG= COPTS.generic="-Os -fomit-frame-pointer -mgnu"
121
Willy Tarreau663148c2012-12-12 00:38:22 +0100122And on a classic Linux with SSL and ZLIB support (eg: Red Hat 5.x) :
123
124 $ make TARGET=linux26 CPU=native USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1
125
126And on a recent Linux >= 2.6.28 with SSL and ZLIB support :
Willy Tarreaud4508812012-09-10 09:07:41 +0200127
William Lallemand82fe75c2012-10-23 10:25:10 +0200128 $ make TARGET=linux2628 CPU=native USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1
Willy Tarreaud4508812012-09-10 09:07:41 +0200129
William Lallemand82fe75c2012-10-23 10:25:10 +0200130In order to build a 32-bit binary on an x86_64 Linux system with SSL support
131without support for compression but when OpenSSL requires ZLIB anyway :
Willy Tarreauef7341d2009-04-11 19:45:50 +0200132
Willy Tarreaud4508812012-09-10 09:07:41 +0200133 $ make TARGET=linux26 ARCH=i386 USE_OPENSSL=1 ADDLIB=-lz
Willy Tarreauef7341d2009-04-11 19:45:50 +0200134
Willy Tarreau3b8e9792012-11-22 00:43:09 +0100135The BSD and OSX makefiles do not support build options for OpenSSL nor zlib.
136Also, at least on OpenBSD, pthread_mutexattr_setpshared() does not exist so
137the SSL session cache cannot be shared between multiple processes. If you want
138to enable these options, you need to use GNU make with the default makefile as
139follows :
140
141 $ gmake TARGET=openbsd USE_OPENSSL=1 USE_ZLIB=1 USE_PRIVATE_CACHE=1
142
willy tarreau78345332005-12-18 01:33:16 +0100143If you need to pass other defines, includes, libraries, etc... then please
144check the Makefile to see which ones will be available in your case, and
Willy Tarreau83b30c12008-05-25 10:32:50 +0200145use the USE_* variables in the GNU Makefile, or ADDINC, ADDLIB, and DEFINE
146variables in the BSD makefiles.
willy tarreau78345332005-12-18 01:33:16 +0100147
Willy Tarreau97ec9692010-01-28 20:52:05 +0100148AIX 5.3 is known to work with the generic target. However, for the binary to
149also run on 5.2 or earlier, you need to build with DEFINE="-D_MSGQSUPPORT",
150otherwise __fd_select() will be used while not being present in the libc.
Willy Tarreau7dec9652012-06-06 16:15:03 +0200151If you get build errors because of strange symbols or section mismatches,
152simply remove -g from DEBUG_CFLAGS.
Willy Tarreau97ec9692010-01-28 20:52:05 +0100153
Willy Tarreau32e65ef2013-04-02 08:14:29 +0200154You can easily define your own target with the GNU Makefile. Unknown targets
155are processed with no default option except USE_POLL=default. So you can very
156well use that property to define your own set of options. USE_POLL can even be
157disabled by setting USE_POLL="". For example :
158
159 $ gmake TARGET=tiny USE_POLL="" TARGET_CFLAGS=-fomit-frame-pointer
160
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200161
1622) How to install it
163--------------------
164
165To install haproxy, you can either copy the single resulting binary to the
166place you want, or run :
167
168 $ sudo make install
169
170If you're packaging it for another system, you can specify its root directory
171in the usual DESTDIR variable.
172
173
1743) How to set it up
175-------------------
176
177There is some documentation in the doc/ directory :
178
179 - architecture.txt : this is the architecture manual. It is quite old and
180 does not tell about the nice new features, but it's still a good starting
181 point when you know what you want but don't know how to do it.
182
183 - configuration.txt : this is the configuration manual. It recalls a few
184 essential HTTP basic concepts, and details all the configuration file
185 syntax (keywords, units). It also describes the log and stats format. It
186 is normally always up to date. If you see that something is missing from
Willy Tarreau74774c02014-04-23 00:57:08 +0200187 it, please report it as this is a bug. Please note that this file is
188 huge and that it's generally more convenient to review Cyril Bonté's
189 HTML translation online here :
190
191 http://cbonte.github.io/haproxy-dconv/configuration-1.5.html
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200192
193 - haproxy-en.txt / haproxy-fr.txt : these are the old outdated docs. You
194 should never need them. If you do, then please report what you didn't
195 find in the other ones.
196
197 - gpl.txt / lgpl.txt : the copy of the licenses covering the software. See
198 the 'LICENSE' file at the top for more information.
199
200 - the rest is mainly for developers.
201
202There are also a number of nice configuration examples in the "examples"
203directory as well as on several sites and articles on the net which are linked
204to from the haproxy web site.
205
206
2074) How to report a bug
208----------------------
209
210It is possible that from time to time you'll find a bug. A bug is a case where
211what you see is not what is documented. Otherwise it can be a misdesign. If you
212find that something is stupidly design, please discuss it on the list (see the
213"how to contribute" section below). If you feel like you're proceeding right
214and haproxy doesn't obey, then first ask yourself if it is possible that nobody
215before you has even encountered this issue. If it's unlikely, the you probably
216have an issue in your setup. Just in case of doubt, please consult the mailing
217list archives :
218
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200219 http://marc.info/?l=haproxy
220
221Otherwise, please try to gather the maximum amount of information to help
222reproduce the issue and send that to the mailing list :
223
224 haproxy@formilux.org
225
226Please include your configuration and logs. You can mask your IP addresses and
227passwords, we don't need them. But it's essential that you post your config if
228you want people to guess what is happening.
229
230Also, keep in mind that haproxy is designed to NEVER CRASH. If you see it die
231without any reason, then it definitely is a critical bug that must be reported
232and urgently fixed. It has happened a couple of times in the past, essentially
233on development versions running on new architectures. If you think your setup
234is fairly common, then it is possible that the issue is totally unrelated.
235Anyway, if that happens, feel free to contact me directly, as I will give you
236instructions on how to collect a usable core file, and will probably ask for
237other captures that you'll not want to share with the list.
238
239
2405) How to contribute
241--------------------
242
243It is possible that you'll want to add a specific feature to satisfy your needs
244or one of your customers'. Contributions are welcome, however I'm often very
245picky about changes. I will generally reject patches that change massive parts
246of the code, or that touch the core parts without any good reason if those
247changes have not been discussed first.
248
249The proper place to discuss your changes is the HAProxy Mailing List. There are
250enough skilled readers to catch hazardous mistakes and to suggest improvements.
Willy Tarreau9a639a12011-09-10 22:48:36 +0200251I trust a number of them enough to merge a patch if they say it's OK, so using
252the list is the fastest way to get your code reviewed and merged. You can
253subscribe to it by sending an empty e-mail at the following address :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200254
255 haproxy+subscribe@formilux.org
256
Willy Tarreau9a639a12011-09-10 22:48:36 +0200257If you have an idea about something to implement, *please* discuss it on the
258list first. It has already happened several times that two persons did the same
259thing simultaneously. This is a waste of time for both of them. It's also very
260common to see some changes rejected because they're done in a way that will
261conflict with future evolutions, or that does not leave a good feeling. It's
262always unpleasant for the person who did the work, and it is unpleasant for me
263too because I value people's time and efforts. That would not happen if these
264were discussed first. There is no problem posting work in progress to the list,
265it happens quite often in fact. Also, don't waste your time with the doc when
266submitting patches for review, only add the doc with the patch you consider
267ready to merge.
268
Willy Tarreau2ddccb72013-05-01 10:07:21 +0200269Another important point concerns code portability. Haproxy requires gcc as the
270C compiler, and may or may not work with other compilers. However it's known
271to build using gcc 2.95 or any later version. As such, it is important to keep
272in mind that certain facilities offered by recent versions must not be used in
273the code :
274
275 - declarations mixed in the code (requires gcc >= 3.x)
276 - GCC builtins without checking for their availability based on version and
277 architecture ;
278 - assembly code without any alternate portable form for other platforms
279 - use of stdbool.h, "bool", "false", "true" : simply use "int", "0", "1"
280 - in general, anything which requires C99 (such as declaring variables in
281 "for" statements)
282
283Since most of these restrictions are just a matter of coding style, it is
284normally not a problem to comply.
285
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200286If your work is very confidential and you can't publicly discuss it, you can
287also mail me directly about it, but your mail may be waiting several days in
288the queue before you get a response.
289
290If you'd like a feature to be added but you think you don't have the skills to
291implement it yourself, you should follow these steps :
292
293 1. discuss the feature on the mailing list. It is possible that someone
294 else has already implemented it, or that someone will tell you how to
295 proceed without it, or even why not to do it. It is also possible that
296 in fact it's quite easy to implement and people will guide you through
297 the process. That way you'll finally have YOUR patch merged, providing
298 the feature YOU need.
299
300 2. if you really can't code it yourself after discussing it, then you may
301 consider contacting someone to do the job for you. Some people on the
302 list might be OK with trying to do it. Otherwise, you can check the list
303 of contributors at the URL below, some of the regular contributors may
304 be able to do the work, probably not for free but their time is as much
305 valuable as yours after all, you can't eat the cake and have it too.
306
307The list of past and regular contributors is available below. It lists not only
308significant code contributions (features, fixes), but also time or money
309donations :
310
311 http://haproxy.1wt.eu/contrib.html
312
313Note to contributors: it's very handy when patches comes with a properly
Willy Tarreau9a639a12011-09-10 22:48:36 +0200314formated subject. There are 3 criteria of particular importance in any patch :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200315
Willy Tarreau9a639a12011-09-10 22:48:36 +0200316 - its nature (is it a fix for a bug, a new feature, an optimization, ...)
317 - its importance, which generally reflects the risk of merging/not merging it
318 - what area it applies to (eg: http, stats, startup, config, doc, ...)
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200319
Willy Tarreau9a639a12011-09-10 22:48:36 +0200320It's important to make these 3 criteria easy to spot in the patch's subject,
321because it's the first (and sometimes the only) thing which is read when
322reviewing patches to find which ones need to be backported to older versions.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200323
Willy Tarreau9a639a12011-09-10 22:48:36 +0200324Specifically, bugs must be clearly easy to spot so that they're never missed.
325Any patch fixing a bug must have the "BUG" tag in its subject. Most common
326patch types include :
327
328 - BUG fix for a bug. The severity of the bug should also be indicated
329 when known. Similarly, if a backport is needed to older versions,
330 it should be indicated on the last line of the commit message. If
331 the bug has been identified as a regression brought by a specific
332 patch or version, this indication will be appreciated too. New
333 maintenance releases are generally emitted when a few of these
334 patches are merged.
335
336 - CLEANUP code cleanup, silence of warnings, etc... theorically no impact.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200337 These patches will rarely be seen in stable branches, though they
Willy Tarreau9a639a12011-09-10 22:48:36 +0200338 may appear when they remove some annoyance or when they make
339 backporting easier. By nature, a cleanup is always minor.
340
341 - REORG code reorganization. Some blocks may be moved to other places,
342 some important checks might be swapped, etc... These changes
343 always present a risk of regression. For this reason, they should
344 never be mixed with any bug fix nor functional change. Code is
345 only moved as-is. Indicating the risk of breakage is highly
346 recommended.
347
348 - BUILD updates or fixes for build issues. Changes to makefiles also fall
349 into this category. The risk of breakage should be indicated if
350 known. It is also appreciated to indicate what platforms and/or
351 configurations were tested after the change.
352
353 - OPTIM some code was optimised. Sometimes if the regression risk is very
354 low and the gains significant, such patches may be merged in the
355 stable branch. Depending on the amount of code changed or replaced
356 and the level of trust the author has in the change, the risk of
357 regression should be indicated.
358
359 - RELEASE release of a new version (development or stable).
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200360
Willy Tarreau9a639a12011-09-10 22:48:36 +0200361 - LICENSE licensing updates (may impact distro packagers).
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200362
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200363
Willy Tarreau9a639a12011-09-10 22:48:36 +0200364When the patch cannot be categorized, it's best not to put any tag. This is
365commonly the case for new features, which development versions are mostly made
366of.
367
368Additionally, the importance of the patch should be indicated when known. A
369single upper-case word is preferred, among :
370
371 - MINOR minor change, very low risk of impact. It is often the case for
372 code additions that don't touch live code. For a bug, it generally
373 indicates an annoyance, nothing more.
374
375 - MEDIUM medium risk, may cause unexpected regressions of low importance or
376 which may quickly be discovered. For a bug, it generally indicates
377 something odd which requires changing the configuration in an
378 undesired way to work around the issue.
379
380 - MAJOR major risk of hidden regression. This happens when I rearrange
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200381 large parts of code, when I play with timeouts, with variable
382 initializations, etc... We should only exceptionally find such
Willy Tarreau9a639a12011-09-10 22:48:36 +0200383 patches in stable branches. For a bug, it indicates severe
384 reliability issues for which workarounds are identified with or
385 without performance impacts.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200386
Willy Tarreau9a639a12011-09-10 22:48:36 +0200387 - CRITICAL medium-term reliability or security is at risk and workarounds,
388 if they exist, might not always be acceptable. An upgrade is
389 absolutely required. A maintenance release may be emitted even if
390 only one of these bugs are fixed. Note that this tag is only used
391 with bugs. Such patches must indicate what is the first version
392 affected, and if known, the commit ID which introduced the issue.
393
394If this criterion doesn't apply, it's best not to put it. For instance, most
395doc updates and most examples or test files are just added or updated without
396any need to qualify a level of importance.
397
398The area the patch applies to is quite important, because some areas are known
399to be similar in older versions, suggesting a backport might be desirable, and
400conversely, some areas are known to be specific to one version. When the tag is
401used alone, uppercase is preferred for readability, otherwise lowercase is fine
402too. The following tags are suggested but not limitative :
403
404 - doc documentation updates or fixes. No code is affected, no need to
405 upgrade. These patches can also be sent right after a new feature,
406 to document it.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200407
Willy Tarreau9a639a12011-09-10 22:48:36 +0200408 - examples example files. Be careful, sometimes these files are packaged.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200409
Willy Tarreau9a639a12011-09-10 22:48:36 +0200410 - tests regression test files. No code is affected, no need to upgrade.
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200411
Willy Tarreau9a639a12011-09-10 22:48:36 +0200412 - init initialization code, arguments parsing, etc...
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200413
Willy Tarreau9a639a12011-09-10 22:48:36 +0200414 - config configuration parser, mostly used when adding new config keywords
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200415
Willy Tarreau9a639a12011-09-10 22:48:36 +0200416 - http the HTTP engine
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200417
Willy Tarreau9a639a12011-09-10 22:48:36 +0200418 - stats the stats reporting engine as well as the stats socket CLI
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200419
Willy Tarreau9a639a12011-09-10 22:48:36 +0200420 - checks the health checks engine (eg: when adding new checks)
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200421
Willy Tarreau9a639a12011-09-10 22:48:36 +0200422 - acl the ACL processing core or some ACLs from other areas
423
424 - peers the peer synchronization engine
425
426 - listeners everything related to incoming connection settings
427
428 - frontend everything related to incoming connection processing
429
430 - backend everything related to LB algorithms and server farm
431
432 - session session processing and flags (very sensible, be careful)
433
434 - server server connection management, queueing
435
436 - proxy proxy maintenance (start/stop)
437
438 - log log management
439
440 - poll any of the pollers
441
442 - halog the halog sub-component in the contrib directory
443
444 - contrib any addition to the contrib directory
445
446Other names may be invented when more precise indications are meaningful, for
447instance : "cookie" which indicates cookie processing in the HTTP core. Last,
448indicating the name of the affected file is also a good way to quickly spot
449changes. Many commits were already tagged with "stream_sock" or "cfgparse" for
450instance.
451
452It is desired that AT LEAST one of the 3 criteria tags is reported in the patch
453subject. Ideally, we would have the 3 most often. The two first criteria should
454be present before a first colon (':'). If both are present, then they should be
455delimited with a slash ('/'). The 3rd criterion (area) should appear next, also
456followed by a colon. Thus, all of the following messages are valid :
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200457
458Examples of messages :
Willy Tarreau9a639a12011-09-10 22:48:36 +0200459 - DOC: document options forwardfor to logasap
460 - DOC/MAJOR: reorganize the whole document and change indenting
461 - BUG: stats: connection reset counters must be plain ascii, not HTML
462 - BUG/MINOR: stats: connection reset counters must be plain ascii, not HTML
463 - MEDIUM: checks: support multi-packet health check responses
464 - RELEASE: Released version 1.4.2
465 - BUILD: stats: stdint is not present on solaris
466 - OPTIM/MINOR: halog: make fgets parse more bytes by blocks
467 - REORG/MEDIUM: move syscall redefinition to specific places
468
469Please do not use square brackets anymore around the tags, because they give me
470more work when merging patches. By default I'm asking Git to keep them but this
471causes trouble when patches are prefixed with the [PATCH] tag because in order
472not to store it, I have to hand-edit the patches. So as of now, I will ask Git
473to remove whatever is located between square brackets, which implies that any
474subject formatted the old way will have its tag stripped out.
475
476In fact, one of the only square bracket tags that still makes sense is '[RFC]'
477at the beginning of the subject, when you're asking for someone to review your
478change before getting it merged. If the patch is OK to be merged, then I can
479merge it as-is and the '[RFC]' tag will automatically be removed. If you don't
480want it to be merged at all, you can simply state it in the message, or use an
481alternate '[WIP]' tag ("work in progress").
482
483The tags are not rigid, follow your intuition first, anyway I reserve the right
484to change them when merging the patch. It may happen that a same patch has a
485different tag in two distinct branches. The reason is that a bug in one branch
486may just be a cleanup in the other one because the code cannot be triggered.
487
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200488
489For a more efficient interaction between the mainline code and your code, I can
490only strongly encourage you to try the Git version control system :
491
492 http://git-scm.com/
493
494It's very fast, lightweight and lets you undo/redo your work as often as you
495want, without making your mistakes visible to the rest of the world. It will
496definitely help you contribute quality code and take other people's feedback
497in consideration. In order to clone the HAProxy Git repository :
498
499 $ git clone http://git.1wt.eu/git/haproxy-1.4.git (stable 1.4)
500 $ git clone http://git.1wt.eu/git/haproxy.git/ (development)
501
Willy Tarreau663148c2012-12-12 00:38:22 +0100502The site above is slow, a faster mirror is maintained up to date here :
503
504 $ git clone http://master.formilux.org/git/people/willy/haproxy.git/
505
Willy Tarreaub1a34b62010-05-09 22:37:12 +0200506If you decide to use Git for your developments, then your commit messages will
507have the subject line in the format described above, then the whole description
508of your work (mainly why you did it) will be in the body. You can directly send
509your commits to the mailing list, the format is convenient to read and process.
510
willy tarreau78345332005-12-18 01:33:16 +0100511-- end