In TCP, we don't want to forward chunks of data, we want to forward
indefinitely. This patch introduces a special value for the amount
of data to be forwarded. When buffer_forward() is called with
BUF_INFINITE_FORWARD, it configures the buffer to never stop
forwarding until the end.
[BUG] stream_interface: SI_ST_CLO must have buffers SHUT
An abort during a connect would go to the SI_ST_CLO state without
the buffers shut. This was causing some sessions to never end if
they would abort before the connect request was initiated. This
bug has been introduced after 1.4-dev2.
The BF_EMPTY flag was once used to indicate an empty buffer. However,
it was used half the time as meaning the buffer is empty for the reader,
and half the time as meaning there is nothing left to send.
"nothing to send" is only indicated by "->send_max=0 && !pipe". Once
we fix this, we discover that the flag is not used anymore. So the
flags has been renamed BF_OUT_EMPTY and means exactly the condition
above, ie, there is nothing to send.
Doing so has allowed us to remove some unused tests for emptiness,
but also to uncover a certain amount of situations where the flag
was not correctly set or tested.
[MAJOR] buffers: split BF_WRITE_ENA into BF_AUTO_CONNECT and BF_AUTO_CLOSE
The BF_WRITE_ENA buffer flag became very complex to deal with, because
it was used to :
- enable automatic connection
- enable close forwarding
- enable data forwarding
The last point was not very true anymore since we introduced ->send_max,
but still the test remained everywhere. This was causing issues such as
impossibility to connect without forwarding data, impossibility to prevent
closing when data was forwarded, etc...
This patch clarifies the situation by getting rid of this multi-purpose
flag and replacing it with :
- data forwarding based only on ->send_max || ->pipe ;
- a new BF_AUTO_CONNECT flag to allow automatic connection and only
that ;
- ability to perform an automatic connection when ->send_max or ->pipe
indicate that data is waiting to leave the buffer ;
- a new BF_AUTO_CLOSE flag to let the producer automatically set the
BF_SHUTW_NOW flag when it gets a BF_SHUTR.
During this cleanup, it was discovered that some tests were performed
twice, or that the BF_HIJACK flag was still tested, which is not needed
anymore since ->send_max replcaed it. These places have been fixed too.
These cleanups have also revealed a few areas where the other flags
such as BF_EMPTY are not cleanly used. This will be an opportunity for
a second patch.
By inlining this function and slightly reordering it, we can double
the getchar/putchar test throughput, and reduce its footprint by about
40 bytes. Also, it was the only non-inlined char-based function, which
now makes it more consistent this time.
[MAJOR] http: add support for HTTP 1xx informational responses
HTTP supports status codes 100 and 101 to report protocol indications,
which are followed by the requests's response. Till now, haproxy would
only see those responses without parsing subsequent ones. That means
that cookie additions were only performed on 1xx messages for instance,
which does not work since headers must be ignored with 1xx messages.
Also, logs were not terribly useful with the common 100 status code
in response to "Expect: 100-continue" during POST some requests.
This change adds support for such messages. Now haproxy sees them,
forwards them and skips them until it finds a correct response, which
it logs and processes. As an exception, header removal/rewriting still
work on 1xx responses in order to be able to strip out sensible
information that may have accidentely been left by another equipment
(possibly an older haproxy itself). But headers addition are disabled
however.
This change brings the ability to loop on response without data, which
is a starting point to support keepalive. The change is marked as major
as a few fixes had to be performed in the HTTP message parser.
[MEDIUM] stream_interface: make use of buffer_cut_tail() to report errors
The stream_int_return() function used to call buffer_erase() on the response
buffer, which completely wipes it without taking care about whatever could
have been there. Now we more carefully strip only data not scheduled to be
sent.
[MINOR] buffers: add buffer_cut_tail() to cut only unsent data
This function is used to cut the "tail" of a buffer, which means strip it
to the length of unsent data only, and kill any remaining unsent data. Any
scheduled forwarding is stopped. This is mainly to be used to send error
messages after existing data. It does the same as buffer_erase() for buffers
without pending outgoing data.
[BUG] buffer_forward() would not correctly consider data already scheduled
The computations in buffer_forward() were only valid if buffer_forward()
was used on a buffer which had no more data scheduled for forwarding.
This is always the case right now so this bug is not yet triggered but
it will soon be. Now we correctly discount the bytes to be forwarded
from the data already present in the buffer.
[MINOR] buffers: add peekchar and peekline functions for stream interfaces
The buffer_si_peekline() function is sort of a fgets() to be used from a
stream interface. It returns a complete line whenever possible, and does
not update the buffer's pointer, so that the reader is free to consume
what it wants to.
buffer_si_peekchar() only returns one character, and also needs a call
to buffer_skip() once the character is definitely consumed.
Willy Tarreau [Mon, 31 Aug 2009 06:09:57 +0000 (08:09 +0200)]
[MEDIUM] buffers: provide new buffer_feed*() function
This functions act like their buffer_write*() counter-parts,
except that they're specifically designed to be used from a
stream interface handler, as they carefully check size limits
and automatically advance the read pointer depending on the
to_forward attribute.
buffer_feed_chunk() is an inline calling buffer_feed() as both
are the sames. For this reason, buffer_write_chunk() has also
been turned into an inline which calls buffer_write().
Willy Tarreau [Mon, 31 Aug 2009 05:37:22 +0000 (07:37 +0200)]
[MINOR] buffers: provide more functions to handle buffer data
buffer_contig_space(), buffer_contig_data() and buffer_skip()
provide easy methods to extract/insert data from/into a buffer.
buffer_write() and buffer_write_chunk() currently do not check
max_len nor to_forward, so they will quickly become embarrassing
to use or will need an equivalent. The reason is that they are
used to build error messages which currently are not subject to
analysis.
[MAJOR] buffers: fix misuse of the BF_SHUTW_NOW flag
This flag was incorrectly used as meaning "close immediately",
while it needs to say "close ASAP". ASAP here means when unsent
data pending in the buffer are sent. This helps cleaning up some
dirty tricks where the buffer output was checking the BF_SHUTR
flag combined with EMPTY and other such things. Now we have a
clearly defined semantics :
- producer sets SHUTR and *may* set SHUTW_NOW if WRITE_ENA is
set, otherwise leave it to the session processor to set it.
- consumer only checks SHUTW_NOW to decide whether or not to
call shutw().
This also induced very minor changes at some locations which were
not protected against buffer changes while the SHUTW_NOW flag was
set. Now we prevent send_max from changing when the flag is set.
Several tests have been run without any unexpected behaviour detected.
Some more cleanups are needed, as it clearly appears that some tests
could be removed with stricter semantics.
Willy Tarreau [Sun, 30 Aug 2009 22:17:18 +0000 (00:17 +0200)]
[BUG] tarpit did not work anymore
Tarpit was broken by recent splitting of analysers. It would still
let the connection go to the server due to a missing buffer_write_dis().
Also, it was performed too late (after content switching rules).
Willy Tarreau [Wed, 19 Aug 2009 09:14:11 +0000 (11:14 +0200)]
[MEDIUM] remove TCP_CORK and make use of MSG_MORE instead
send() supports the MSG_MORE flag on Linux, which does the same
as TCP_CORK except that we don't have to remove TCP_NODELAY before
and we don't need any syscall to set/remove it. This can save up
to 4 syscalls around a send() (two for setting it, two for removing
it), and it's much cleaner since it is not persistent. So make use
of it instead.
Willy Tarreau [Tue, 18 Aug 2009 05:38:19 +0000 (07:38 +0200)]
[MEDIUM] stream_interface: add and use ->update function to resync
We used to call stream_sock_data_finish() directly at the end of
a session update, but if we want to support non-socket interfaces,
we need to have this function configurable. Now we access it via
->update().
Willy Tarreau [Tue, 18 Aug 2009 05:19:39 +0000 (07:19 +0200)]
[MEDIUM] report error on buffer writes larger than buffer size
Since it's now possible to change the buffer size by configuration,
we have to take special measures against writes that are larger than
the buffer size. Before this patch, the writers would indefinitely
block, waiting for some space to free up.
With this patch, the functions simply reject the data with an
appropriate code so that the writers can either detect and process
the error or go on, but never remain blocked.
This has been tested on the stats page which does no longer hang
with buffer sizes smaller than 2.5 kB (256 bytes is even OK for
the CSV version).
Willy Tarreau [Mon, 17 Aug 2009 05:23:33 +0000 (07:23 +0200)]
[MEDIUM] make it possible to change the buffer size in the configuration
The new tune.bufsize and tune.maxrewrite global directives allow one to
change the buffer size and the maxrewrite size. Right now, setting bufsize
too low will block stats sockets which will not be able to write at all.
An error checking must be added to buffer_write_chunk() so that if it
cannot write its message to an empty buffer, it causes the caller to abort.
Willy Tarreau [Sun, 16 Aug 2009 21:27:46 +0000 (23:27 +0200)]
[MEDIUM] replace BUFSIZE with buf->size in computations
The first step towards dynamic buffer size consists in removing
all static definitions of the buffer size. Instead, we store a
buffer's size in itself. Right now they're all preinitialized
to BUFSIZE, but we will change that.
Willy Tarreau [Sun, 16 Aug 2009 20:57:50 +0000 (22:57 +0200)]
[MEDIUM] set rep->analysers from fe and be analysers
sess_establish() used to resort to protocol-specific guesses
in order to set rep->analysers. This is no longer needed as it
gets set from the frontend and the backend as a copy of what
was defined in the configuration.
Willy Tarreau [Sun, 16 Aug 2009 20:45:38 +0000 (22:45 +0200)]
[MINOR] set s->srv_error according to the analysers
s->srv_error was set depending on the frontend's protocol. Now it is
set by the HTTP analyser, so that even when switching from a TCP
frontend to an HTTP backend, we can have HTTP error messages.
Willy Tarreau [Sun, 16 Aug 2009 20:37:44 +0000 (22:37 +0200)]
[MINOR] cleanup set_session_backend by using pre-computed analysers
Analyser bitmaps are now stored in the frontend and backend, and
combined at configuration time. That way, set_session_backend()
does not need to perform any protocol-specific combinations.
Willy Tarreau [Sun, 16 Aug 2009 17:12:36 +0000 (19:12 +0200)]
[MINOR] move the initial task's nice value to the listener
Since the listener is the one indicating what analyser and session
handlers to call, it makes sense that it also sets the task's nice
value. This also helps getting rid of the last trace of the stats
in the proto_uxst file.
Willy Tarreau [Sun, 16 Aug 2009 17:06:42 +0000 (19:06 +0200)]
[CLEANUP] move remaining stats sockets code to dumpstats
The remains of the stats socket code has nothing to do in proto_uxst
anymore and must move to dumpstats. The code is much cleaner and more
structured. It was also an opportunity to rename AN_REQ_UNIX_STATS
as AN_REQ_STATS_SOCK as the stats socket is no longer unix-specific
either.
The last item refering to stats in proto_uxst is the setting of the
task's nice value which should in fact come from the listener.
Willy Tarreau [Sun, 16 Aug 2009 16:51:29 +0000 (18:51 +0200)]
[MEDIUM] make the unix stats sockets use the generic session handler
process_session() is now ready to handle unix stats sockets. This
first step works and old code has not been removed. A cleanup is
required. The stats handler is not unix socket-centric anymore and
should move to dumpstats.c.
Willy Tarreau [Sun, 16 Aug 2009 16:27:24 +0000 (18:27 +0200)]
[MINOR] session: switch to established state if no connect function
When a stream interface has no connect() function, it means it is
immediately connected, so we don't need any connection request.
This will be used with unix sockets.
In order to merge the unix session handling code, we have to maintain
the number of per-listener connections in the session. This was only
performed for unix sockets till now.
Willy Tarreau [Sun, 16 Aug 2009 15:41:45 +0000 (17:41 +0200)]
[MEDIUM] make the global stats socket part of a frontend
Creating a frontend for the global stats socket will help merge
unix sockets management with the other socket management. Since
frontends are huge structs, we only allocate it if required.
Willy Tarreau [Sun, 16 Aug 2009 12:02:45 +0000 (14:02 +0200)]
[MEDIUM] move connection establishment from backend to the SI.
The connection establishment was completely handled by backend.c which
normally just handles LB algos. Since it's purely TCP, it must move to
proto_tcp.c. Also, instead of calling it directly, we now call it via
the stream interface, which will later help us unify session handling.
Willy Tarreau [Sun, 16 Aug 2009 12:13:47 +0000 (14:13 +0200)]
[BUILD] fix build for systems without SOL_TCP
Andrew Azarov reported that haproxy-1.4-dev1 does not build
under FreeBSD 7.2 because SOL_TCP is not defined. So add a
check for its definition before using it. This only impacts
network optimisations anyway.
Willy Tarreau [Sun, 16 Aug 2009 11:20:32 +0000 (13:20 +0200)]
[MEDIUM] remove old experimental tcpsplice option
This Linux-specific option was never really used in production and
has since been superseded by new splicing options brought by recent
Linux kernels.
It caused several particular cases in the code because the kernel
would take care of the session without haproxy being able to do
anything on it, which became hard to handle in the new architecture.
Let's simply get rid of it now that there is a replacement available.
Willy Tarreau [Sun, 16 Aug 2009 08:29:18 +0000 (10:29 +0200)]
[MINOR] stats: add a new node-name setting
The new "node-name" stats setting enables reporting of a node ID on
the stats page. It is possible to return the system's host name as
well as a specific name.
Willy Tarreau [Sun, 9 Aug 2009 20:57:09 +0000 (22:57 +0200)]
[RELEASE] Released version 1.4-dev2
Released version 1.4-dev2 with the following main changes :
- [BUG] task: fix possible crash when some timeouts are not configured
- [BUG] log: option tcplog would log to global if no logger was defined
Willy Tarreau [Sun, 9 Aug 2009 08:11:45 +0000 (10:11 +0200)]
[BUG] log: option tcplog would log to global if no logger was defined
Romuald du Song reported a strange bug causing "option tcplog" to
unexpectedly use global log parameters if no log server was declared.
Eventhough it can be useful in some circumstances, it only hides
configuration bugs and can even cause traffic logs to be sent to
the wrong logger, since global settings are just for the process.
This has been fixed and a warning has been added for configurations
where tcplog or httplog are set without any logger. This fix must
be backported to 1.3.20, but not to 1.3.15.X in order not to risk
any regression on old configurations.
Willy Tarreau [Sun, 9 Aug 2009 07:09:54 +0000 (09:09 +0200)]
[BUG] task: fix possible crash when some timeouts are not configured
Cristian Ditoiu reported a major regression when testing 1.3.19 at
transfer.ro. It would crash within a few minutes while 1.3.15.10
was OK. He offered to help so we could run gdb and debug the crash
live. We finally found that the crash was the result of a regression
introduced by recent fix 814c978fb67782ceeaf1db74abfe7083938bedff
(task: fix possible timer drift after update) which makes it possible
for a tree walk to start from a detached task if this task has got
its timeout disabled due to a missing timeout.
The trivial fix below has been extensively tested and confirmed not
to crash anymore.
Special thanks to Cristian who spontaneously provided a lot of help
and trust to debug this issue which at first glance looked impossible
after reading the code and traces, but took less than an hour to spot
and fix when caught live in gdb ! That's really appreciated !
Released version 1.4-dev1 with the following main changes :
- [MINOR] acl: add support for matching of RDP cookies
- [MEDIUM] add support for RDP cookie load-balancing
- [MEDIUM] add support for RDP cookie persistence
- [MINOR] add a new CLF log format
- [MINOR] startup: don't imply -q with -D
- [BUG] ensure that we correctly re-start old process in case of error
- [MEDIUM] add support for binding to source port ranges during connect
- [MINOR] config: track "no option"/"option" changes
- [MINOR] config: support resetting options do default values
- [MEDIUM] implement option tcp-smart-accept at the frontend
- [MEDIUM] stream_sock: implement tcp-cork for use during shutdowns on Linux
- [MEDIUM] implement tcp-smart-connect option at the backend
- [MEDIUM] add support for TCP MSS adjustment for listeners
- [MEDIUM] support setting a server weight to zero
- [MINOR] make DEFAULT_MAXCONN user-configurable at build time
- [MAJOR] session: don't clear buffer status flags anymore
- [MAJOR] session: only check for timeouts when they have just occurred.
- [MAJOR] session: simplify buffer error handling
- [MEDIUM] config: split parser and checker in two functions
- [MEDIUM] config: support loading multiple configuration files
- [MEDIUM] stream_sock: don't close prematurely when nolinger is set
- [MEDIUM] session: rework buffer analysis to permit permanent analysers
- [MEDIUM] splice: set the capability on each stream_interface
- [BUG] http: redirect rules were processed too early
- [CLEANUP] remove unused DEBUG_PARSE_NO_SPEEDUP define
- [MEDIUM] http: split request waiter from request processor
- [MEDIUM] session: tell analysers what bit they were called for
- [MAJOR] http: complete splitting of the remaining stages
- [MINOR] report in the proxies the requirements for ACLs
- [MINOR] http: rely on proxy->acl_requires to allocate hdr_idx
- [MINOR] acl: add HTTP protocol detection (req_proto_http)
- [MINOR] prepare callers of session_set_backend to handle errors
- [BUG] default ACLs did not properly set the ->requires flag
- [MEDIUM] allow a TCP frontend to switch to an HTTP backend
- [MINOR] ensure we can jump from swiching rules to http without data
- [MINOR] http: take http request timeout from the backend
- [MINOR] allow TCP inspection rules to make use of HTTP ACLs
- [BUILD] report commit date and not author's date as build date
- [MINOR] acl: don't complain anymore when using L7 acls in TCP
- [BUG] stream_sock: always shutdown(SHUT_WR) before closing
- [BUG] stream_sock: don't stop reading when the poller reports an error
- [BUG] config: tcp-request content only accepts "if" or "unless"
- [BUG] task: fix possible timer drift after update
- [MINOR] apply tcp-smart-connect option for the checks too
- [MINOR] stats: better displaying in MSIE
- [MINOR] config: improve error reporting in global section
- [MINOR] config: improve error reporting in listen sections
- [MINOR] config: the "capture" keyword is not allowed in backends
- [MINOR] config: improve error reporting when checking configuration
- [BUILD] fix a minor build warning on AIX
- [BUILD] use "git cmd" instead of "git-cmd"
- [CLEANUP] report 2009 not 2008 in the copyright banner.
- [MINOR] print usage on the stats sockets upon invalid commands
- [MINOR] acl: detect and report potential mistakes in ACLs
- [BUILD] fix incorrect printf arg count with tcp_splice
- [BUG] fix random pauses on last segment of a series
- [BUILD] add support for build under Cygwin
[BUG] fix random pauses on last segment of a series
During a direct data transfer from the server to the client, if the
system did not have enough buffers anymore, haproxy would not enable
write polling again if it could write at least one data chunk. Under
normal conditions, this would remain undetected because the remaining
data would be pushed by next data chunks.
However, when this happens on the last chunk of a session, or the last
in a series in an interactive bidirectional TCP transfer, haproxy would
only start sending again when the read timeout was reached on the side
it stopped writing, causing long pauses on some protocols such as SQL.
This bug was reported by an Exceliance customer who generously offered
to help us by sending large amounts of traces and running various tests
on production systems.
It is quite hard to trigger it but it becomes easier with a ping-pong
TCP service which transfers random data sizes, with a modified version
of send() able to send packets smaller than the average transfer size.
A cleaner fix would imply only updating the write timeout when data
transfers are *attempted*, not succeeded, but that requires more
sensible code changes without fixing the result. It is a candidate
for a later patch though.
[MINOR] acl: detect and report potential mistakes in ACLs
I've discovered a configuration with lots of occurrences of the
following :
acl xxx hdr_beg (host) xxx
The problem is that hdr_beg will match every header against patterns
(host) and xxx due to the space between both, which certainly is not
what the user wanted. Now we detect such ACLs and report a warning
with a suggestion to add "--" between "hdr_beg" and "(host)" if this
is definitely what is wanted.
[MINOR] print usage on the stats sockets upon invalid commands
When issuing commands on the unix socket, there's no way to
know if the result is empty or if the command is wrong. This
patch makes invalid command return a help message.
Newer GIT versions do not support "git-cmd" anymore, so date and version
can be wrong during development builds. Use "git cmd" now. Also fix
git-tar to use "git archive" instead of "git-tar-tree".
[MINOR] config: improve error reporting when checking configuration
Do not exit early at the first error found while checking configuration
validity. This particularly helps spotting multiple wrong tracked server
names at once.
[MINOR] config: improve error reporting in listen sections
Try not to immediately exit on non-fatal errors while parsing a
listen section, so that the user has a chance to get most of the
errors at once, which is quite convenient especially during config
checks with the -c argument.
[MINOR] config: improve error reporting in global section
Try not to immediately exit on non-fatal errors while parsing the
global section, so that the user has a chance to get most of the
errors at once, which is quite convenient especially during config
checks with the -c argument. Some other errors such as unresolved
server names also don't make the parser exit too early.
MSIE does not correctly display spaced digits. It requires a margin of
at least one pixel. Also, it does not correctly hide empty cells, so we
work around this by setting the background white. Last, the H1 font was
too large, so we reduce it by one size, which is still OK in other
browsers.
[MINOR] apply tcp-smart-connect option for the checks too
We should respect tcp-smart-connect for checks too. First it reduces
the traffic, and second it ensures that the checks see the same thing
as the production traffic, which is better for debugging.
When the scheduler detected that a task was misplaced in the timer
queue, it used to place it right again. Unfortunately, it did not
check whether it would still call the new task from its new place.
This resulted in some tasks not getting called on timeout once in
a while, causing a minor drift for repetitive timers. This effect
was only observable with slow health checks and without any activity
because no other task would cause the scheduler to be immediately
called again.
In practice, it does not affect any real-world configuration, but
it's still better to fix it.
[BUG] config: tcp-request content only accepts "if" or "unless"
As reported by Maik Broemme, if something different from "if" or
"unless" was specified after "tcp-request content accept", the
condition would silently remain void. The parser must obviously
complain since this typically corresponds to a forgotten "if".
[BUG] stream_sock: don't stop reading when the poller reports an error
As reported by Jean-Baptiste Quenot and Robbie Aelter, sometimes a
backend server error is converted to a 502 error if the backend stops
before reading all the request. The reason is that the remote system
sends a TCP RST packet because there are still unread data pending in
the socket buffer. This RST is translated as a socket error on the
local system, and this error is reported by the poller.
However, most of the time, it's a write error, but the system is
still able to read the remaining pending data, such as in the trace
below :
The recv succeeded while epoll_wait() reported an error.
Note: This case is very hard to reproduce and requires that the backend
server is reached via the loopback in order to minimise latency and
reduce the risk of sent data being ACKed.
[BUG] stream_sock: always shutdown(SHUT_WR) before closing
When we close a socket with unread data in the buffer, or when the
nolinger option is set, we regularly lose the last fragment, which
often contains the error message. This typically occurs when sending
too large a request. Only the RST is seen due to the close() (since
not all data were read) and the output message never reaches the
network.
Doing a shutdown() before the close() solves this annoying issue
because the data are really pushed before the system sends the RST.
[BUILD] report commit date and not author's date as build date
By default, when building from a git tree, haproxy's release date is
set to the last commit's date. But it was the wrong date which was
used, the initial patch's date, which can cause time jumps in the
past when an old patch gets merged. What we want is the commit date,
which reflects the correct code history.
Emeric Brun [Tue, 30 Jun 2009 15:57:00 +0000 (17:57 +0200)]
[MEDIUM] add support for RDP cookie persistence
The new statement "persist rdp-cookie" enables RDP cookie
persistence. The RDP cookie is then extracted from the RDP
protocol, and compared against available servers. If a server
matches the RDP cookie, then it gets the connection.
Emeric Brun [Tue, 30 Jun 2009 15:56:00 +0000 (17:56 +0200)]
[MEDIUM] add support for RDP cookie load-balancing
This patch adds support for hashing RDP cookies in order to
use them as a load-balancing key. The new "rdp-cookie(name)"
load-balancing metric has to be used for this. It is still
mandatory to wait for an RDP cookie in the frontend, otherwise
it will randomly work.
Emeric Brun [Tue, 30 Jun 2009 15:54:00 +0000 (17:54 +0200)]
[MINOR] acl: add support for matching of RDP cookies
The RDP protocol is quite simple and documented, which permits
an easy detection and extraction of cookies. It can be useful
to match the MSTS cookie which can contain the username specified
by the client.
[MINOR] allow TCP inspection rules to make use of HTTP ACLs
Since we can call the HTTP parser from TCP inspection rules, it makes
sense to be able to use the HTTP ACLs with it. That way, we can decide
from a TCP frontend to take a switching decision based on full layer7
decoding. This might be useful to perform layer7 content switching from
a layer4 frontend in fact. For instance, we might want to be able to
detect http/https on a frontend, but still switch to backend X or Y
depending on the Host header. Note that it is mandatory to wait for
an HTTP request otherwise the ACLs will randomly match.
[MINOR] http: take http request timeout from the backend
Since we can now switch from TCP to HTTP, we need to be able to apply
the HTTP request timeout after switching. That means we need to take
it from the backend and not from the frontend. Since the backend points
to the frontend before switching, that changes nothing for the normal
case.
[MINOR] ensure we can jump from swiching rules to http without data
In case of switching from TCP to HTTP, we want the HTTP request timeout
to be properly initialized. For this, we have to jump to the analyser
without breaking out of the loop nor waiting for incoming data. The way
it is done right now is not particularly clean but it works.
A cleaner method might involve pushing function pointers into a circular
list.
[MEDIUM] allow a TCP frontend to switch to an HTTP backend
This patch allows a TCP frontend to switch to an HTTP backend.
During the switch, missing structures are automatically allocated.
The HTTP parser is enabled so that the backend first waits for a
full HTTP request.
Now that we can perform TCP-based content switching, it makes sense
to be able to detect HTTP traffic and act accordingly. We already
have an HTTP decoder, we just have to call it in order to detect HTTP
protocol. Note that since the decoder will automatically fill in the
interesting fields of the HTTP transaction, it would make sense to
use this parsing to extend HTTP matching to TCP.
[MINOR] http: rely on proxy->acl_requires to allocate hdr_idx
Right now only HTTP proxies may use HTTP headers in ACLs, but
when this evolves, we'll need to be able to allocate the hdr_idx
on demand. The solution consists in allocating it only when it is
certain that at least one ACL requires HTTP parsing, regardless
of the mode the proxy is in. This is what is achieved by this
patch.
[MINOR] report in the proxies the requirements for ACLs
This patch propagates the ACL conditions' "requires" bitfield
to the proxies. This makes it possible to know exactly what a
proxy might have to support for any request, which helps knowing
whether we have to allocate some space for certain types of
structures or not (eg: the hdr_idx struct).
The concept might be extended to a lot more types of information,
such as detecting whether we need to allocate some space for some
request ACLs which need a result in the response, etc...
[MAJOR] http: complete splitting of the remaining stages
The HTTP processing has been splitted into 7 steps, one of which
is not anymore HTTP-specific (content-switching). That way, it
becomes possible to use "use_backend" rules in TCP mode. A new
"use_server" directive should follow soon.
[MEDIUM] session: tell analysers what bit they were called for
Some stream analysers might become generic enough to be called
for several bits. So we cannot have the analyser bit hard coded
into the analyser itself. Let's make the caller inform the callee.
[MEDIUM] http: split request waiter from request processor
We want to split several steps in HTTP processing so that
we can call individual analysers depending on what processing
we want to perform. The first step consists in splitting the
part that waits for a request from the rest.
[BUG] http: redirect rules were processed too early
redirect rules are documented as being processed last before
use_backend but were mistakenly processed before block rules.
Fortunately very few people use a mix of block and redirect
rules, so this bug has never been reported yet.
Willy Tarreau [Sun, 28 Jun 2009 21:10:19 +0000 (23:10 +0200)]
[MEDIUM] splice: set the capability on each stream_interface
The splice code did not consider compatibility between both ends
of the connection. Now we set different capabilities on each
stream interface, depending on what the protocol can splice to/from.
Right now, only TCP is supported. Thanks to this, we're now able to
automatically detect when splice() is not implemented and automatically
disable it on one end instead of reporting errors to the upper layer.
Willy Tarreau [Sun, 28 Jun 2009 17:37:53 +0000 (19:37 +0200)]
[MEDIUM] session: rework buffer analysis to permit permanent analysers
It will soon be necessary to support permanent analysers (eg: HTTP in
keep-alive mode). We first have to slightly rework the call to the
request analysers so that we don't force ->analysers to be 0 before
forwarding data.
Willy Tarreau [Sun, 28 Jun 2009 09:09:07 +0000 (11:09 +0200)]
[MEDIUM] stream_sock: don't close prematurely when nolinger is set
When the nolinger option is used, we must not close too fast because
some data might be left unsent. Instead we must proceed with a normal
shutdown first, then a close. Also, we want to avoid merging FIN with
the last segment if nolinger is set, because if that one gets lost,
there is no chance for it to be retransmitted.
Willy Tarreau [Mon, 22 Jun 2009 13:48:36 +0000 (15:48 +0200)]
[MEDIUM] config: split parser and checker in two functions
This is a first step towards support of multiple configuration files.
Now readcfgfile() only reads a file in memory and performs very minimal
parsing. The checks are performed afterwards.
Willy Tarreau [Sun, 21 Jun 2009 20:43:05 +0000 (22:43 +0200)]
[MAJOR] session: simplify buffer error handling
Buffer errors (timeouts and I/O errors) were handled at two places,
just after the analysers and after again.
Now that the timeout detection has moved, it has become easier to
handle those errors.
This has also made it possible for the request and response analysers
to be processed together as a down-up event, and all the up-down I/O
updates to be processed afterwards, which is exactly what we're looking
for. Interestingly this has reduced the number of iterations of
(stream_int, req_resp) from (5,6,5) to (5,5,4).
Several tests have been run without any issue found.
Willy Tarreau [Sun, 21 Jun 2009 20:03:51 +0000 (22:03 +0200)]
[MAJOR] session: only check for timeouts when they have just occurred.
It's useless to check for buffer timeouts every time we call
process_session() because we already control when we set the flag. So
let's check them at the precise moment where the flag is set.
Willy Tarreau [Sun, 21 Jun 2009 19:45:58 +0000 (21:45 +0200)]
[MAJOR] session: don't clear buffer status flags anymore
We want to be able to keep information about errors and timeouts
as long as possible in the buffer. Let's not clear these flags
anymore and keep them static. This does not seem to cause any
trouble, though a finer review might be wise.
Willy Tarreau [Sun, 14 Jun 2009 16:48:19 +0000 (18:48 +0200)]
[MEDIUM] add support for TCP MSS adjustment for listeners
Sometimes it can be useful to limit the advertised TCP MSS on
incoming connections, for instance when requests come through
a VPN or when the system is running with jumbo frames enabled.
Passing the "mss <value>" arguments to a "bind" line will set
the value. This works under Linux >= 2.6.28, and maybe a few
earlier ones, though due to an old kernel bug most of earlier
versions will probably ignore it. It is also possible that some
other OSes will support this.
Yitzhak Sapir [Sun, 14 Jun 2009 16:27:54 +0000 (18:27 +0200)]
[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).
Willy Tarreau [Sun, 14 Jun 2009 13:48:17 +0000 (15:48 +0200)]
[MEDIUM] implement tcp-smart-connect option at the backend
This new option enables combining of request buffer data with
the initial ACK of an outgoing TCP connection. Doing so saves
one packet per connection which is quite noticeable on workloads
mostly consisting in small objects. The option is not enabled by
default.
Willy Tarreau [Sun, 14 Jun 2009 13:24:37 +0000 (15:24 +0200)]
[MEDIUM] stream_sock: implement tcp-cork for use during shutdowns on Linux
Setting TCP_CORK on a socket before sending the last segment enables
automatic merging of this segment with the FIN from the shutdown()
call. Playing with TCP_CORK is not easy though as we have to track
the status of the TCP_NODELAY flag since both are mutually exclusive.
Doing so saves one more packet per session and offers about 5% more
performance.
There is no reason not to do it, so there is no associated option.
Willy Tarreau [Sun, 14 Jun 2009 10:07:01 +0000 (12:07 +0200)]
[MEDIUM] implement option tcp-smart-accept at the frontend
This option disables TCP quick ack upon accept. It is also
automatically enabled in HTTP mode, unless the option is
explicitly disabled with "no option tcp-smart-accept".
This saves one packet per connection which can bring reasonable
amounts of bandwidth for servers processing small requests.
Willy Tarreau [Sun, 14 Jun 2009 09:39:52 +0000 (11:39 +0200)]
[MINOR] config: support resetting options do default values
A new keyword prefix "default" has been introduced in order to
reset some options to their default values. This can be needed
for instance when an option is forced disabled or enabled in a
defaults section and when later sections want to use automatic
settings regardless of what was specified there. Right now it
is only supported by options, just like the "no" prefix.
Willy Tarreau [Sun, 14 Jun 2009 09:10:45 +0000 (11:10 +0200)]
[MINOR] config: track "no option"/"option" changes
Sometimes we would want to implement implicit default options,
but for this we need to be able to disable them, which requires
to keep track of "no option" settings. With this change, an option
explicitly disabled in a defaults section will still be seen as
explicitly disabled. There should be no regression as nothing makes
use of this yet.
Willy Tarreau [Wed, 10 Jun 2009 09:09:37 +0000 (11:09 +0200)]
[MEDIUM] add support for binding to source port ranges during connect
Some users are already hitting the 64k source port limit when
connecting to servers. The system usually maintains a list of
unused source ports, regardless of the source IP they're bound
to. So in order to go beyond the 64k concurrent connections, we
have to manage the source ip:port lists ourselves.
The solution consists in assigning a source port range to each
server and use a free port in that range when connecting to that
server, either for a proxied connection or for a health check.
The port must then be put back into the server's range when the
connection is closed.
This mechanism is used only when a port range is specified on
a server. It makes it possible to reach 64k connections per
server, possibly all from the same IP address. Right now it
should be more than enough even for huge deployments.
Willy Tarreau [Tue, 9 Jun 2009 12:36:00 +0000 (14:36 +0200)]
[BUG] ensure that we correctly re-start old process in case of error
When a new process fails to grab some ports, it sends a signal to
the old process in order to release them. Then it tries to bind
again. If it still fails (eg: one of the ports is bound to a
completely different process), it must send the continue signal
to the old process so that this one re-binds to the ports. This
is correctly done, but the newly bound ports are not released
first, which sometimes causes the old process to remain running
with no port bound. The fix simply consists in unbinding all
ports before sending the signal to the old process.
Willy Tarreau [Mon, 18 May 2009 14:29:51 +0000 (16:29 +0200)]
[MINOR] startup: don't imply -q with -D
It is recommended to have -D in init scripts, but -D also implies
quiet mode, which hides warning messages, and both options are now
completely unrelated. Remove the implication to get warnings with
-D.