Commit Graph

11 Commits

Author SHA1 Message Date
Willy Tarreau 60612ebbbf [RELEASE] Released version 1.5-dev7
Released version 1.5-dev7 with the following main changes :
    - [BUG] fix binary stick-tables
    - [MINOR] http: *_dom matching header functions now also split on ":"
    - [BUG] checks: fix support of Mysqld >= 5.5 for mysql-check
    - [MINOR] acl: add srv_conn acl to count connections on a     specific backend server
    - [MINOR] check: add redis check support
    - [DOC] small fixes to clearly distinguish between keyword     and variables
    - [MINOR] halog: add support for termination code matching (-tcn/-TCN)
    - [DOC] Minor spelling fixes and grammatical enhancements
    - [CLEANUP] dumpstats: make symbols static where possible
    - [MINOR] Break out dumping table
    - [MINOR] Break out processing of clear table
    - [MINOR] Allow listing of stick table by key
    - [MINOR] Break out all stick table socat command parsing
    - [MINOR] More flexible clearing of stick table
    - [MINOR] Allow showing and clearing by key of ipv6 stick tables
    - [MINOR] Allow showing and clearing by key of integer stick tables
    - [MINOR] Allow showing and clearing by key of string stick tables
    - [CLEANUP] Remove assigned but unused variables
    - [CLEANUP] peers.h: fix declarations
    - [CLEANUP] session.c: Make functions static where possible
    - [MINOR] Add active connection list to server
    - [MINOR] Allow shutdown of sessions when a server becomes unavailable
    - [MINOR] Add down termination condition
    - [MINOR] Make appsess{,ion}_refresh static
    - [MINOR] Add rdp_cookie pattern fetch function
    - [CLEANUP] Remove unnecessary casts
    - [MINOR] Add non-stick server option
    - [MINOR] Consistently use error in tcp_parse_tcp_req()
    - [MINOR] Consistently free expr on error in cfg_parse_listen()
    - [MINOR] Free rdp_cookie_name on denint()
    - [MINOR] Free tcp rules on denint()
    - [MINOR] Free stick table pool on denint()
    - [MINOR] Free stick rules on denint()
    - [MEDIUM] Fix stick-table replication on soft-restart
    - [MEDIUM] Correct ipmask() logic
    - [MINOR] Correct type in table dump examples
    - [MINOR] Fix build error in stream_int_register_handler()
    - [MINOR] Use DPRINTF in assign_server()
    - [BUG] checks: http-check expect could fail a check on multi-packet responses
    - [DOC] fix minor typo in the "dispatch" doc
    - [BUG] proto_tcp: fix address binding on remote source
    - [MINOR] http: don't report the "haproxy" word on the monitoring response
    - [REORG] http: move HTTP error codes back to proto_http.h
    - [MINOR] http: make the "HTTP 200" status code configurable.
    - [MINOR] http: partially revert the chunking optimization for now
    - [MINOR] stream_sock: always clear BF_EXPECT_MORE upon complete transfer
    - [CLEANUP] stream_sock: remove unneeded FL_TCP and factor out test
    - [MEDIUM] http: add support for "http-no-delay"
    - [OPTIM] http: optimize chunking again in non-interactive mode
    - [OPTIM] stream_sock: avoid fast-forwarding of partial data
    - [OPTIM] stream_sock: don't use splice on too small payloads
    - [MINOR] config: make it possible to specify a cookie even without a server
    - [BUG] stats: support url-encoded forms
    - [MINOR] config: automatically compute a default fullconn value
    - [CLEANUP] config: remove some left-over printf debugging code from previous patch
    - [DOC] add missing entry or stick store-response
    - [MEDIUM] http: add support for 'cookie' and 'set-cookie' patterns
    - [BUG] halog: correctly handle truncated last line
    - [MINOR] halog: make SKIP_CHAR stop on field delimiters
    - [MINOR] halog: add support for HTTP log matching (-H)
    - [MINOR] halog: gain back performance before SKIP_CHAR fix
    - [OPTIM] halog: cache some common fields positions
    - [OPTIM] halog: check once for correct line format and reuse the pointer
    - [OPTIM] halog: remove many 'if' by using a function pointer for the filters
    - [OPTIM] halog: remove support for tab delimiters in input data
    - [BUG] session: risk of crash on out of memory (1.5-dev regression)
    - [MINOR] session: try to emit a 500 response on memory allocation errors
    - [OPTIM] stream_sock: reduce the default number of accepted connections at once
    - [BUG] stream_sock: disable listener when system resources are exhausted
    - [MEDIUM] proxy: add a PAUSED state to listeners and move socket tricks out of proxy.c
    - [BUG] stream_sock: ensure orphan listeners don't accept too many connections
    - [MINOR] listeners: add listen_full() to mark a listener full
    - [MINOR] listeners: add support for queueing resource limited listeners
    - [MEDIUM] listeners: put listeners in queue upon resource shortage
    - [MEDIUM] listeners: queue proxy-bound listeners at the proxy's
    - [MEDIUM] listeners: don't stop proxies when global maxconn is reached
    - [MEDIUM] listeners: don't change listeners states anymore in maintain_proxies
    - [CLEANUP] proxy: rename a few proxy states (PR_STIDLE and PR_STRUN)
    - [MINOR] stats: report a "WAITING" state for sockets waiting for resource
    - [MINOR] proxy: make session rate-limit more accurate
    - [MINOR] sessions: only wake waiting listeners up if rate limit is OK
    - [BUG] proxy: peers must only be stopped once, not upon every call to maintain_proxies
    - [CLEANUP] proxy: merge maintain_proxies() operation inside a single loop
    - [MINOR] task: new function task_schedule() to schedule a wake up
    - [MAJOR] proxy: finally get rid of maintain_proxies()
    - [BUG] proxy: stats frontend and peers were missing many initializers
    - [MEDIUM] listeners: add a global listener management task
    - [MINOR] proxy: make findproxy() return proxies from numeric IDs too
    - [DOC] fix typos, "#" is a sharp, not a dash
    - [MEDIUM] stats: add support for changing frontend's maxconn at runtime
    - [MEDIUM] checks: group health checks methods by values and save option bits
    - [MINOR] session-counters: add the ability to clear the counters
    - [BUG] check: http-check expect + regex would crash in defaults section
    - [MEDIUM] http: make x-forwarded-for addition conditional
    - [REORG] build: move syscall redefinition to specific places
    - [CLEANUP] update the year in the copyright banner
    - [BUG] possible crash in 'show table' on stats socket
    - [BUG] checks: use the correct destination port for sending checks
    - [BUG] backend: risk of picking a wrong port when mapping is used with crossed families
    - [MINOR] make use of set_host_port() and get_host_port() to get rid of family mismatches
    - [DOC] fixed a few "sensible" -> "sensitive" errors
    - [MINOR] make use of addr_to_str() and get_host_port() to replace many inet_ntop()
    - [BUG] http: trailing white spaces must also be trimmed after headers
    - [MINOR] stats: display "<NONE>" instead of the frontend name when unknown
    - [MINOR] http: take a capture of too large requests and responses
    - [MINOR] http: take a capture of truncated responses
    - [MINOR] http: take a capture of bad content-lengths.
    - [DOC] add a few old and uncommitted docs
    - [CLEANUP] cfgparse: fix reported options for the "bind" keyword
    - [MINOR] halog: add -hs/-HS to filter by HTTP status code range
    - [MINOR] halog: support backslash-escaped quotes
    - [CLEANUP] remove dirty left-over of a debugging message
    - [MEDIUM] stats: disable complex socket reservation for stats socket
    - [CLEANUP] remove a useless test in manage_global_listener_queue()
    - [MEDIUM] stats: add the "set maxconn" setting to the command line interface
    - [MEDIUM] add support for global.maxconnrate to limit the per-process conn rate.
    - [MINOR] stats: report the current and max global connection rates
    - [MEDIUM] stats: add the ability to adjust the global maxconnrate
    - [BUG] peers: don't pre-allocate 65000 connections to each peer
    - [MEDIUM] don't limit peers nor stats socket to maxconn nor maxconnrate
    - [BUG] peers: the peer frontend must not emit any log
    - [CLEANUP] proxy: make pause_proxy() perform the required controls and emit the logs
    - [BUG] peers: don't keep a peers section which has a NULL frontend
    - [BUG] peers: ensure the peers are resumed if they were paused
    - [MEDIUM] stats: add the ability to enable/disable/shutdown a frontend at runtime
    - [MEDIUM] session: make session_shutdown() an independant function
    - [MEDIUM] stats: offer the possibility to kill a session from the CLI
    - [CLEANUP] stats: centralize tests for backend/server inputs on the CLI
    - [MEDIUM] stats: offer the possibility to kill sessions by server
    - [MINOR] halog: do not consider byte 0x8A as end of line
    - [MINOR] frontend: ensure debug message length is always initialized
    - [OPTIM] halog: make fgets parse more bytes by blocks
    - [OPTIM] halog: add assembly version of the field lookup code
    - [MEDIUM] poll: add a measurement of idle vs work time
    - [CLEANUP] startup: report only the basename in the usage message
    - [MINOR] startup: add an option to change to a new directory
    - [OPTIM] task: don't scan the run queue if we know it's empty
    - [BUILD] stats: stdint is not present on solaris
    - [DOC] update the README file to reflect new naming rules for patches
    - [MINOR] stats: report the number of requests intercepted by the frontend
    - [DOC] update ROADMAP file
2011-09-10 23:43:11 +02:00
Willy Tarreau 9a639a13cd [DOC] update the README file to reflect new naming rules for patches
Those will be in effect after 1.5-dev7 is released.
2011-09-10 22:48:36 +02:00
Willy Tarreau 64bc40b654 [BUILD] add the USE_GETADDRINFO build option
This one is used to call getaddrinfo() to resolve IPv6 host names.
2011-03-23 22:49:56 +01:00
Willy Tarreau a5899aaad5 [BUILD] add the CPU=native and ARCH=32/64 build options
Hank A. Paulson suggested to add CPU=native to optimize the code for
the build machine. This makes sense in a lot of situations. Since it
is often possible to have both 32 and 64 bits supported on recent
systems, the ARCH=32 and ARCH=64 build options were also added.
2010-11-28 07:49:59 +01:00
Willy Tarreau b1a34b68ca [DOC] refresh the README file and merge the CONTRIB file into it
Patrick Mézard reported that it was a bit awkward to have the CONTRIB
and contrib entries in the source archive since those can conflict on
case-insensitive file systems. That made a good opportunity to refresh
the README file and to remove that old outdated file.
2010-05-09 22:37:12 +02:00
Willy Tarreau 97ec969077 [DOC] add some build info about the AIX platform
(cherry picked from commit e41914c77e)
2010-01-29 10:30:19 +01:00
Yitzhak Sapir 32087312e3 [BUILD] add support for build under Cygwin
After considering various possibilities, we compiled haproxy under cygwin.
Attached is an updated full diff that also has the TARGET=cygwin documented.
The whole thing compiles and installs with this diff only.

In cygwin 1.7 (now in beta), there is apparently support for ipv6.  Cygwin
1.5 (later versions, anyway) already includes some support in the form of a
define USE_IPV6.  When defined, it declares the sockaddr_in6 struct and
possibly other things.  The above definition AF_INET6=23 is taken from
their /usr/include/socket.h file (where it is #if 0'd out).

We are running into a socket limit.  It appears that Cygwin (running on
Windows 2003 Server) will only allow us to set ulimit -n (maximum open
files) to 3200, which means we're a little short of 1600 connections.

The limit of 3200 is an internal Cygwin limit.  Perhaps they can raise it in
the future.  Using the nbproc option, I was able to bring up 10 servers.  It
seems to me that they were able to handle over 2000 connections (even though
each had maxconn 1500 set, and the hard Cygwin fd limit).
2009-06-14 18:27:54 +02:00
Willy Tarreau ef7341dc3d [BUILD] make it possible to pass alternative arch at build time
When trying to build a 32-bit binary on a 64-bit platform, we generally
need to pass "-m32" to gcc, which is not convenient with current makefile.
Note that this option requires gcc >= 3.

In order to ease parameter passing, a new ARCH= makefile option has been
added. If it receives a target architecture, according "-m32"/"-m64" and
"-march=xxxx" will be passed to gcc. Only the generic makefile has been
changed to support this option right now as the need only appeared on Linux.

The spec file now makes use of this option so that rpmbuild can automatically
build with the proper architecture.
2009-04-11 20:32:54 +02:00
Willy Tarreau 83b30c1e3c [DOC] update the README file with new build options 2008-05-25 10:32:50 +02:00
willy tarreau d38e72d567 Separated OpenBSD build from the main Makefile into a new one. 2006-03-19 20:56:52 +01:00
willy tarreau 783453346f * added a README to help about the build process 2005-12-18 01:33:16 +01:00