The accept loop used to force fd_poll_recv() even in places where it
was not completely appropriate (eg: unexpected errors). It does not
yet cause trouble but will do with the upcoming polling changes. Let's
use it only where relevant now. EINTR/ECONNABORTED do not result in
poll() anymore but the failed connection is simply skipped (this code
dates from 1.1.32 when error codes were first considered).
Some rare unexplained busy loops were observed on versions up to 1.5-dev19.
It happens that if a file descriptor happens to be disabled for both read and
write while it was speculatively enabled for both and this without creating a
new update entry, there will be no way to remove it from the speculative I/O
list until some other changes occur. It is suspected that a double sequence
such as enable_both/disable_both could have led to this situation where an
update cancels itself and does not clear the spec list in the poll loop.
While it is unclear what I/O sequence may cause this situation to arise, it
is safer to always add the FD to the update list if nothing could be done on
it so that the next poll round will automatically take care of it.
This is 1.5-specific, no backport is needed.
Recent commit abf08d9 ("BUG/MAJOR: connection: fix mismatch between rcv_buf's
API and usage") accidentely broke SSL by relying on an uninitialized value to
enter the read loop.
Many thanks to Cyril Bonté and Steve Ruiz for reporting this issue.
The value of the variable "appctx->ctx.map.ent" is used after the loop,
but its value has changed. The variable "value" is initialized and
contains the good value.
This is a recent bug, no backport is needed.
Recent commit 4448925 ("BUILD/MINOR: listener: remove a glibc warning on accept4()")
broke accept4() on some systems because the glibc's version may now conflict with
the local one.
Steve Ruiz reported some reproducible crashes with HTTP health checks
on a certain page returning a huge length. The traces he provided
clearly showed that the recv() call was performed twice for a total
size exceeding the buffer's length.
Cyril Bonté tracked down the problem to be caused by the full buffer
size being passed to rcv_buf() in event_srv_chk_r() instead of passing
just the remaining amount of space. Indeed, this change happened during
the connection rework in 1.5-dev13 with the following commit :
f150317 MAJOR: checks: completely use the connection transport layer
But one of the problems is also that the comments at the top of the
rcv_buf() functions suggest that the caller only has to ensure the
requested size doesn't overflow the buffer's size.
Also, these functions already have to care about the buffer's size to
handle wrapping free space when there are pending data in the buffer.
So let's change the API instead to more closely match what could be
expected from these functions :
- the caller asks for the maximum amount of bytes it wants to read ;
This means that only the caller is responsible for enforcing the
reserve if it wants to (eg: checks don't).
- the rcv_buf() functions fix their computations to always consider
this size as a max, and always perform validity checks based on
the buffer's free space.
As a result, the code is simplified and reduced, and made more robust
for callers which now just have to care about whether they want the
buffer to be filled or not.
Since the bug was introduced in 1.5-dev13, no backport to stable versions
is needed.
The accept4() Linux syscall requires _GNU_SOURCE on ix86, otherwise
it emits a warning. On other archs including x86_64, this problem
doesn't happen. Thanks to Charles Carter from Sigma Software for
reporting this.
If the pattern is set as case insensitive, the string comparison
is executed twice. The first time is insensitive comparison, the
second is sensitive.
This is a recent bug, no backport is needed.
This reverts commit 1208266356.
It randomly breaks SSL. What happens is that if the SSL response is
read at once by the SSL stack and is partially delivered to the buffer,
then there's no way to read the next parts because we wait for some
polling first.
So we'll fix this after the polling rework.
A config where multiple servers have the same name in the same backend is
prone to a number of issues : logs are not really exploitable, stats get
really tricky and even harder to change, etc...
In fact, it can be safe to have the same name between multiple servers only
when their respective IDs are known and used. So now we detect this situation
and emit a warning for the first conflict detected per server if any of the
servers uses an automatic ID.
The code which enables tunnel mode or TCP transfers is rarely used
and at most once per session. Putting it in an unlikely() clause
reduces the length of the hot path of process_session() which is
already quite long, and also slightly reduces its overall size.
Some measurements show a steady gain of about 0.2% thanks to this.
This function is called twice per request, and does almost always nothing.
Better use an inline version to avoid entering it when we can.
About 0.5% additional performance was gained this way.
si_connect() used to only return SI_ST_CON. But it already detect the
connection reuse and is the function which avoids calling connect().
So it already knows the connection is valid and reuse. Thus we make it
return SI_ST_EST when a connection is reused. This means that
connect_server() can return this state and sess_update_stream_int()
as well.
Thanks to this change, we don't need to leave process_session() in
SI_ST_CON state to immediately enter it again to switch to SI_ST_EST.
Implementing this removes one call to process_session() per request
in keep-alive mode. We're now at 2 calls per request, which is the
minimum (one for the request and another one for the response). The
number of calls to http_wait_for_response() has also dropped from 2
to one.
Tests indicate a performance gain of about 2.6% in request rate in
keep-alive mode. There should be no gain in http-server-close() since
we don't use this faster path.
At the moment it is possible in sess_prepare_conn_req() to switch to the
established state when the target is an applet. But sess_update_stream_int()
will soon also have the ability to set the established state via
connect_server() when a connection is reused, leading to a synchronous
connect.
So prepare the code to handle this SI_ST_ASS -> SI_ST_EST transition, which
really matches what's done in the lower layers.
Currently there are 3 places in the code where t_connect is set after
switching to state SI_ST_EST, and a fourth one will soon come. Since
all these places lead to an immediate call to sess_establish() to
complete the session establishment, better move that measurement
there.
It's a bit hasardous to wipe out all channel flags, this flag should
be left intact as it protects against recursive calls. Fortunately,
we have no possibility to meet this situation with current applets,
but better fix it before it becomes an issue.
This bug has been there for a long time, but it doesn't seem worth
backporting the fix.
As soon as we connect to the server, we want to limit the number of
recvfrom() on the response path because most of the time a single
call will retrieve enough information.
At the moment this is only done in the HTTP response parser, after
some reads have already failed, which is too late. We need to do
that at the earliest possible instant. It was already done for the
request side by frontend_accept() for the first request, and by
http_reset_txn() for the next requests.
Thanks to this change, there are no more failed recvfrom() calls in
keep-alive mode.
Since commit 6b66f3e ([MAJOR] implement autonomous inter-socket forwarding)
introduced in 1.3.16-rc1, we've been relying on a stupid mechanism to wake
up the task after a write, which was an exact copy-paste of the reader side.
The principle was that if we empty a buffer and there's no forwarding
scheduled or if the *producer* is not in a connected state, then we wake
the task up.
That does not make any sense. It happens to wake up too late sometimes (eg,
when the request analyser waits for some room in the buffer to start to
work), and leads to unneeded wakeups in client-side keep-alive, because
the task is woken up when the response is sent, while the analysers are
simply waiting for a new request.
In order to fix this, we introduce a new channel flag : CF_WAKE_WRITE. It
is designed so that an analyser can explicitly request being notified when
some data were written. It is used only when the HTTP request or response
analysers need to wait for more room in the buffers. It is automatically
cleared upon wake up.
The flag is also automatically set by the functions which try to write into
a buffer from an applet when they fail (bi_putblk() etc...).
That allows us to remove the stupid condition above and avoid some wakeups.
In http-server-close and in http-keep-alive modes, this reduces from 4 to 3
the average number of wakeups per request, and increases the overall
performance by about 1.5%.
This reverts commit f3221f99ac.
Igor reported some very strange breakage of his stats page which is
clearly caused by the chunking, though I don't see at first glance
what could be wrong. Better revert it for now.
In theory the principle is simple as we just need to send HTTP chunks
if the client is 1.1 compatible. In practice it's harder because we
have to append a CR LF after each block of data and we're never sure
to have the room for this. In order not to have to deal with this, we
instead send the CR LF prior to each chunk size. The only issue is for
the first chunk and for this reason we avoid to send the empty header
line when using chunked encoding.
Since the applet rework and the removal of the inter-task applets,
we must not clear the stream-interface's owner task anymore otherwise
we risk a crash when maintaining keep-alive with an applet. This is
not possible right now so there is no impact yet, but this bug is not
easy to track down. No backport is needed.
This value is stored as unsigned in chn->to_forward. Having it defined
as signed makes it impossible to pass channel_forward() a previously
saved value because the argument will be zero-extended during the
conversion to long long, while the test will be performed using sign
extension. There is no impact on existing code right now.
When enabling a tracked server via the web interface, we must first
check if the server tracks another one and the state of this tracked
server, just like the command line does.
Failure to do so causes incorrect logs to be emitted when the server
is enabled :
[WARNING] 361/212556 (2645) : Server bck2/srv3 is DOWN via bck2/srv2. 2 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
[WARNING] 361/212603 (2645) : Server bck2/srv3 is DOWN for maintenance.
--> enable server now
[WARNING] 361/212606 (2645) : Server bck2/srv3 is UP (leaving maintenance).
With this fix, it's correct now :
[WARNING] 361/212805 (2666) : Server bck2/srv3 is DOWN via bck2/srv2. 2 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
[WARNING] 361/212813 (2666) : Server bck2/srv3 is DOWN for maintenance.
--> enable server now
[WARNING] 361/212821 (2666) : Server bck2/srv3 is DOWN via bck2/srv2. 2 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
It does not seem necessary to backport this fix, considering that it
depends on extremely fragile behaviours, there are more risks of breakage
caused by a backport than the current inconvenience.
Recent fix 02541e8 (BUG/MEDIUM: checks: servers must not start in
slowstart mode) failed to consider one case : a server chich is not
checked at all can be disabled and has to support being enabled
again. So we must also enter the set_server_up() function when the
checks are totally disabled.
No backport is needed.
We used to unconditionally disable client-side polling after the client
has posted its request. The goal was to avoid subscribing the file
descriptor to the poller for nothing.
This is perfect for the HTTP close mode where we know we won't have to
read on the client side anymore. However, when keep-alive is maintained
with the client, this makes the situation worse. Indeed, after the first
response, we'll have to wait for the client to send a next request and
since this is never immediate, we'll certainly poll. So what happens is
that polling is enabled after a response and disabled after a request,
so the polling is constantly alternating, which is very expensive with
epoll_ctl().
The solution implemented in this patch consists in only disabling the
polling if the client-side is not in keep-alive mode. That way we have
the best of both worlds. In close, we really close, and in keep-alive,
we poll only once.
The performance gained by this change is important, with haproxy jumping
from 158kreq/s to 184kreq/s (+16%) in HTTP keep-alive mode on a machine
which at best does 222k/s in raw TCP mode.
With this patch and the previous one, a keep-alive run with a fast
enough server (or enough concurrent connections to cover the connect
time) does no epoll_ctl() anymore during a run of ab -k. The net
measured gain is 19%.
If a file descriptor is being polled, and it stopped (eg: buffer full
or end of response), then re-enabled, currently what happens is that
the polling is disabled, then the fd is enabled in speculative mode,
an I/O attempt is made, it loses (otherwise the FD would surely not
have been polled), and the polled is enabled again.
This is too bad, especially with HTTP keep-alive on the server side
where all operations are performed at once before going back to the
poll loop.
Now we improve the behaviour by ensuring that if an fd is still being
polled, when it's enabled after having been disabled, we re-enable the
polling. Doing so saves a number of syscalls and useless wakeups, and
results in a significant performance gain on HTTP keep-alive. A 11%
increase has been observed on the HTTP request rate in keep-alive
thanks to this.
It could be considered as a bug fix, but there was no harm with the
current behaviour, except extra syscalls.
Compression is normally disabled on HTTP/1.0 since it does not
support chunked encoded responses. But the test was incomplete, and
Bertrand Jacquin reported a case where if the server responded using
1.1 to an 1.0 request, then haproxy still used to compress (and of
course the client could not understand the response).
No backport is needed, this is 1.5-specific.
In HTTP keep-alive mode, if we receive a 401, we still have a chance
of being able to send the visitor again to the same server over the
same connection. This is required by some broken protocols such as
NTLM, and anyway whenever there is an opportunity for sending the
challenge to the proper place, it's better to do it (at least it
helps with debugging).
If a server is switched to maintenance mode while a check is in progress,
the successful completion of the check must not switch it back up. This
is still a consequence of using the same function set_server_up() for
every state change. Bug reported by Igor at owind.
This fix should be backported to 1.4 which is affected as well.
This reverts commit 2f877304ef.
This commit is OK for clear text traffic but causes trouble with SSL
when buffers are smaller than SSL buffers. Since the issue it addresses
will be gone once the polling redesign is complete, there's no reason
for trying to workaround temporary inefficiencies. Better remove it.
If we reuse a server-side connection, we must not reinitialize its context nor
try to enable send_proxy. At the moment HTTP keep-alive over SSL fails on the
first attempt because the SSL context was cleared, so it only worked after a
retry.
Commit 2737562 (MEDIUM: stream-int: implement a very simplistic idle
connection manager) implemented an idle connection handler. In the
case where all data is drained from the server, it fails to disable
polling, resulting in a busy spinning loop.
Thanks to Sander Klein and Guillaume Castagnino for reporting this bug.
No backport is needed.
Released version 1.5-dev21 with the following main changes :
- MINOR: stats: don't use a monospace font to report numbers
- MINOR: session: remove debugging code
- BUG/MAJOR: patterns: fix double free caused by loading strings from files
- MEDIUM: http: make option http_proxy automatically rewrite the URL
- BUG/MEDIUM: http: cook_cnt() forgets to set its output type
- BUG/MINOR: stats: correctly report throttle rate of low weight servers
- BUG/MEDIUM: checks: servers must not start in slowstart mode
- BUG/MINOR: acl: parser must also stop at comma on ACL-only keywords
- MEDIUM: stream-int: implement a very simplistic idle connection manager
- DOC: update the ROADMAP file
Idle connections are not monitored right now. So if a server closes after
a response without advertising it, it won't be detected until a next
request wants to use the connection. This is a bit problematic because
it unnecessarily maintains file descriptors and sockets in an idle
state.
This patch implements a very simple idle connection manager for the stream
interface. It presents itself as an I/O callback. The HTTP engine enables
it when it recycles a connection. If a close or an error is detected on the
underlying socket, it tries to drain as much data as possible from the socket,
detect the close and responds with a close as well, then detaches from the
stream interface.
In 1.5-dev20, commit bb9665e (BUG/MEDIUM: checks: ensure we can enable
a server after boot) tried to fix a side effect of having both regular
checks and agent checks condition the up state propagation to servers.
Unfortunately it was still not fine because after this fix, servers
which make use of slowstart start in this mode. We must not check
the agent's health if agent checks are not enabled, and likewise,
we must not check the regular check's health if they are not enabled.
Reading the code, it seems like we could avoid entering this function
at all if (s->state & SRV_RUNNING) is not satisfied. Let's reserve
this for a later patch if needed.
Thanks to Sander Klein for reporting this abnormal situation.
The throttling of low weight servers (<16) could mistakenly be reported
as > 100% due to a rounding that was performed before a multiply by 100
instead of after. This was introduced in 1.5-dev20 when fixing a previous
reporting issue by commit d32c399 (MINOR: stats: report correct throttling
percentage for servers in slowstart).
It should be backported if the patch above is backported.
Since comit b805f71 (MEDIUM: sample: let the cast functions set their
output type), the output type of a fetch function is automatically
considered and passed to the next converter. A bug introduced in
1.5-dev9 with commit f853c46 (MEDIUM: pattern/acl: get rid of
temp_pattern in ACLs) was revealed by this last one : the output type
remained string instead of UINT, causing the cast function to try to
cast the contents and to crash on a NULL deref.
Note: this fix was made after a careful review of all fetch functions.
A few non-trivial ones had their comments amended to clearly indicate
the output type.
There are very few users of http_proxy, and all of them complain about
the same thing : the request is passed unmodified to the server (in its
proxy form), and it is not possible to fix it using reqrep rules because
http_proxy happens after.
So let's have http_proxy fix the URL it has analysed to get rid of the
scheme and the host part. This will do what users of this feature expect.
A null pointer assignment was missing after a free in commit 7148ce6 (MEDIUM:
pattern: Extract the index process from the pat_parse_*() functions), causing
a double free after loading a file of string patterns.
This bug was introduced in 1.5-dev20, no backport is needed.
Thanks to Sander Klein for reporting this bug and providing the config
needed to trigger it.
The memset() was put here to corrupt memory for a debugging test,
it's not needed anymore and was unfortunately committed. It does
not harm anyway, it probably just slightly affects performance.
On several browsers, the monospace font used to display numbers in tips
is not much readable. Since the numbers are aligned anyway, there is too
little benefit in using such a font.
Released version 1.5-dev20 with the following main changes :
- DOC: add missing options to the manpage
- DOC: add manpage references to all system calls
- DOC: update manpage reference to haproxy-en.txt
- DOC: remove -s and -l options from the manpage
- DOC: missing information for the "description" keyword
- DOC: missing http-send-name-header keyword in keyword table
- MINOR: tools: function my_memmem() to lookup binary contents
- MEDIUM: checks: add send/expect tcp based check
- MEDIUM: backend: Enhance hash-type directive with an algorithm options
- MEDIUM: backend: Implement avalanche as a modifier of the hashing functions.
- DOC: Documentation for hashing function, with test results.
- BUG/MEDIUM: ssl: potential memory leak using verifyhost
- BUILD: ssl: compilation issue with openssl v0.9.6.
- BUG/MINOR: ssl: potential memory leaks using ssl_c_key_alg or ssl_c_sig_alg.
- MINOR: ssl: optimization of verifyhost on wildcard certificates.
- BUG/MINOR: ssl: verifyhost does not match empty strings on wildcard.
- MINOR: ssl: Add statement 'verifyhost' to "server" statements
- CLEANUP: session: remove event_accept() which was not used anymore
- BUG/MINOR: deinit: free fdinfo while doing cleanup
- DOC: minor typo fix in documentation
- BUG/MEDIUM: server: set the macro for server's max weight SRV_UWGHT_MAX to SRV_UWGHT_RANGE
- BUG/MINOR: use the same check condition for server as other algorithms
- DOC: fix typo in comments
- BUG/MINOR: deinit: free server map which is allocated in init_server_map()
- CLEANUP: stream_interface: cleanup loop information in si_conn_send_loop()
- MINOR: buffer: align the last output line of buffer_dump()
- MINOR: buffer: align the last output line if there are less than 8 characters left
- DOC: stick-table: modify the description
- OPTIM: stream_interface: return directly if the connection flag CO_FL_ERROR has been set
- CLEANUP: code style: use tabs to indent codes
- DOC: checkcache: block responses with cacheable cookies
- BUG/MINOR: check_config_validity: check the returned value of stktable_init()
- MEDIUM: haproxy-systemd-wrapper: Use haproxy in same directory
- MEDIUM: systemd-wrapper: Kill child processes when interrupted
- LOW: systemd-wrapper: Write debug information to stdout
- BUG/MINOR: http: fix "set-tos" not working in certain configurations
- MEDIUM: http: add IPv6 support for "set-tos"
- DOC: ssl: update build instructions to use new SSL_* variables
- BUILD/MINOR: systemd: fix compiler warning about unused result
- url32+src - like base32+src but whole url including parameters
- BUG/MINOR: fix forcing fastinter in "on-error"
- CLEANUP: Make parameters of srv_downtime and srv_getinter const
- CLEANUP: Remove unused 'last_slowstart_change' field from struct peer
- MEDIUM: Split up struct server's check element
- MEDIUM: Move result element to struct check
- MEDIUM: Paramatise functions over the check of a server
- MEDIUM: cfgparse: Factor out check initialisation
- MEDIUM: Add state to struct check
- MEDIUM: Move health element to struct check
- MEDIUM: Add helper for task creation for checks
- MEDIUM: Add helper function for failed checks
- MEDIUM: Log agent fail, stopped or down as info
- MEDIUM: Remove option lb-agent-chk
- MEDIUM: checks: Add supplementary agent checks
- MEDIUM: Do not mark a server as down if the agent is unavailable
- MEDIUM: Set rise and fall of agent checks to 1
- MEDIUM: Add enable and disable agent unix socket commands
- MEDIUM: Add DRAIN state and report it on the stats page
- BUILD/MINOR: missing header file
- CLEANUP: regex: Create regex_comp function that compiles regex using compilation options
- CLEANUP: The function "regex_exec" needs the string length but in many case they expect null terminated char.
- MINOR: http: some exported functions were not in the header file
- MINOR: http: change url_decode to return the size of the decoded string.
- BUILD/MINOR: missing header file
- BUG/MEDIUM: sample: The function v4tov6 cannot support input and output overlap
- BUG/MINOR: arg: fix error reporting for add-header/set-header sample fetch arguments
- MINOR: sample: export the generic sample conversion parser
- MINOR: sample: export sample_casts
- MEDIUM: acl: use the fetch syntax 'fetch(args),conv(),conv()' into the ACL keyword
- MINOR: stick-table: use smp_expr_output_type() to retrieve the output type of a "struct sample_expr"
- MINOR: sample: provide the original sample_conv descriptor struct to the argument checker function.
- MINOR: tools: Add a function to convert buffer to an ipv6 address
- MINOR: acl: export acl arrays
- MINOR: acl: Extract the pattern parsing and indexation from the "acl_read_patterns_from_file()" function
- MINOR: acl: Extract the pattern matching function
- MINOR: sample: Define new struct sample_storage
- MEDIUM: acl: associate "struct sample_storage" to each "struct acl_pattern"
- REORG: acl/pattern: extract pattern matching from the acl file and create pattern.c
- MEDIUM: pattern: create pattern expression
- MEDIUM: pattern: rename "acl" prefix to "pat"
- MEDIUM: sample: let the cast functions set their output type
- MINOR: sample: add a private field to the struct sample_conv
- MINOR: map: Define map types
- MEDIUM: sample: add the "map" converter
- MEDIUM: http: The redirect strings follows the log format rules.
- BUG/MINOR: acl: acl parser does not recognize empty converter list
- BUG/MINOR: map: The map list was declared in the map.h file
- MINOR: map: Cleanup the initialisation of map descriptors.
- MEDIUM: map: merge identical maps
- BUG/MEDIUM: pattern: Pattern node has type of "struct pat_idx_elt" in place of "struct eb_node"
- BUG/MEDIUM: map: Bad map file parser
- CLEANUP/MINOR: standard: use the system define INET6_ADDRSTRLEN in place of MAX_IP6_LEN
- BUG/MEDIUM: sample: conversion from str to ipv6 may read data past end
- MINOR: map: export map_get_reference() function
- MINOR: pattern: Each pattern sets the expected input type
- MEDIUM: acl: Last patch change the output type
- MEDIUM: pattern: Extract the index process from the pat_parse_*() functions
- MINOR: standard: The function parse_binary() can use preallocated buffer
- MINOR: regex: Change the struct containing regex
- MINOR: regex: Copy the original regex expression into string.
- MINOR: pattern: add support for compiling patterns for lookups
- MINOR: pattern: make the pattern matching function return a pointer to the matched element
- MINOR: map: export parse output sample functions
- MINOR: pattern: add function to lookup a specific entry in pattern list
- MINOR: pattern/map: Each pattern must free the associated sample
- MEDIUM: dumpstat: make the CLI parser understand the backslash as an escape char
- MEDIUM: map: dynamic manipulation of maps
- BUG/MEDIUM: unique_id: junk in log on empty unique_id
- BUG/MINOR: log: junk at the end of syslog packet
- MINOR: Makefile: provide cscope rule
- DOC: compression: chunk are not compressed anymore
- MEDIUM: session: disable lingering on the server when the client aborts
- BUG/MEDIUM: prevent gcc from moving empty keywords lists into BSS
- DOC: remove the comment saying that SSL certs are not checked on the server side
- BUG: counters: third counter was not stored if others unset
- BUG/MAJOR: http: don't emit the send-name-header when no server is available
- BUG/MEDIUM: http: "option checkcache" fails with the no-cache header
- BUG/MAJOR: http: sample prefetch code was not properly migrated
- BUG/MEDIUM: splicing: fix abnormal CPU usage with splicing
- BUG/MINOR: stream_interface: don't call chk_snd() on polled events
- OPTIM: splicing: use splice() for the last block when relevant
- MEDIUM: sample: handle comma-delimited converter list
- MINOR: sample: fix sample_process handling of unstable data
- CLEANUP: acl: move the 3 remaining sample fetches to samples.c
- MINOR: sample: add a new "date" fetch to return the current date
- MINOR: samples: add the http_date([<offset>]) sample converter.
- DOC: minor improvements to the part on the stats socket.
- MEDIUM: sample: systematically pass the keyword pointer to the keyword
- MINOR: payload: split smp_fetch_rdp_cookie()
- MINOR: counters: factor out smp_fetch_sc*_tracked
- MINOR: counters: provide a generic function to retrieve a stkctr for sc* and src.
- MEDIUM: counters: factor out smp_fetch_sc*_get_gpc0
- MEDIUM: counters: factor out smp_fetch_sc*_gpc0_rate
- MEDIUM: counters: factor out smp_fetch_sc*_inc_gpc0
- MEDIUM: counters: factor out smp_fetch_sc*_clr_gpc0
- MEDIUM: counters: factor out smp_fetch_sc*_conn_cnt
- MEDIUM: counters: factor out smp_fetch_sc*_conn_rate
- MEDIUM: counters: factor out smp_fetch_sc*_conn_cur
- MEDIUM: counters: factor out smp_fetch_sc*_sess_cnt
- MEDIUM: counters: factor out smp_fetch_sc*_sess_rate
- MEDIUM: counters: factor out smp_fetch_sc*_http_req_cnt
- MEDIUM: counters: factor out smp_fetch_sc*_http_req_rate
- MEDIUM: counters: factor out smp_fetch_sc*_http_err_cnt
- MEDIUM: counters: factor out smp_fetch_sc*_http_err_rate
- MEDIUM: counters: factor out smp_fetch_sc*_kbytes_in
- MEDIUM: counters: factor out smp_fetch_sc*_bytes_in_rate
- MEDIUM: counters: factor out smp_fetch_sc*_kbytes_out
- MEDIUM: counters: factor out smp_fetch_sc*_bytes_out_rate
- MEDIUM: counters: factor out smp_fetch_sc*_trackers
- MINOR: session: make the number of stick counter entries more configurable
- MEDIUM: counters: support passing the counter number as a fetch argument
- MEDIUM: counters: support looking up a key in an alternate table
- MEDIUM: cli: adjust the method for feeding frequency counters in tables
- MINOR: cli: make it possible to enter multiple values at once with "set table"
- MINOR: payload: allow the payload sample fetches to retrieve arbitrary lengths
- BUG/MINOR: cli: "clear table" must not kill entries that don't match condition
- MINOR: ssl: use MAXPATHLEN instead of PATH_MAX
- MINOR: config: warn when a server with no specific port uses rdp-cookie
- BUG/MEDIUM: unique_id: HTTP request counter must be unique!
- DOC: add a mention about the limited chunk size
- BUG/MEDIUM: fix broken send_proxy on FreeBSD
- MEDIUM: stick-tables: flush old entries upon soft-stop
- MINOR: tcp: add new "close" action for tcp-response
- MINOR: payload: provide the "res.len" fetch method
- BUILD: add SSL_INC/SSL_LIB variables to force the path to openssl
- MINOR: http: compute response time before processing headers
- BUG/MINOR: acl: fix improper string size assignment in proxy argument
- BUG/MEDIUM: http: accept full buffers on smp_prefetch_http
- BUG/MINOR: acl: implicit arguments of ACL keywords were not properly resolved
- BUG/MEDIUM: session: risk of crash on out of memory conditions
- BUG/MINOR: peers: set the accept date in outgoing connections
- BUG/MEDIUM: tcp: do not skip tracking rules on second pass
- BUG/MEDIUM: acl: do not evaluate next terms after a miss
- MINOR: acl: add a warning when an ACL keyword is used without any value
- MINOR: tcp: don't use tick_add_ifset() when timeout is known to be set
- BUG/MINOR: acl: remove patterns from the tree before freeing them
- MEDIUM: backend: add support for the wt6 hash
- OPTIM/MEDIUM: epoll: fuse active events into polled ones during polling changes
- OPTIM/MINOR: mark the source address as already known on accept()
- BUG/MINOR: stats: don't count tarpitted connections twice
- CLEANUP: http: homogenize processing of denied req counter
- CLEANUP: http: merge error handling for req* and http-request *
- BUG/MEDIUM: http: fix possible parser crash when parsing erroneous "http-request redirect" rules
- BUG/MINOR: http: fix build warning introduced with url32/url32_src
- BUG/MEDIUM: checks: fix slow start regression after fix attempt
- BUG/MAJOR: server: weight calculation fails for map-based algorithms
- MINOR: stats: report correct throttling percentage for servers in slowstart
- OPTIM: connection: fold the error handling with handshake handling
- MINOR: peers: accept to learn strings of different lengths
- BUG/MAJOR: fix haproxy crash when using server tracking instead of checks
- BUG/MAJOR: check: fix haproxy crash during soft-stop/soft-start
- BUG/MINOR: stats: do not report "via" on tracking servers in maintenance
- BUG/MINOR: connection: fix typo in error message report
- BUG/MINOR: backend: fix target address retrieval in transparent mode
- BUG/MINOR: config: report the correct track-sc number in tcp-rules
- BUG/MINOR: log: fix log-format parsing errors
- DOC: add some information about how to apply converters to samples
- MINOR: acl/pattern: use types different from int to clarify who does what.
- MINOR: pattern: import acl_find_match_name() into pattern.h
- MEDIUM: stick-tables: support automatic conversion from ipv4<->ipv6
- MEDIUM: log-format: relax parsing of '%' followed by unsupported characters
- BUG/MINOR: http: usual deinit stuff in last commit
- BUILD: log: silent a warning about isblank() with latest patches
- BUG/MEDIUM: checks: fix health check regression causing them to depend on declaration order
- BUG/MEDIUM: checks: fix a long-standing issue with reporting connection errors
- BUG/MINOR: checks: don't consider errno and use conn->err_code
- BUG/MEDIUM: checks: also update the DRAIN state from the web interface
- MINOR: stats: remove some confusion between the DRAIN state and NOLB
- BUG/MINOR: tcp: check that no error is pending during a connect probe
- BUG/MINOR: connection: check EINTR when sending a PROXY header
- MEDIUM: connection: set the socket shutdown flags on socket errors
- BUG/MEDIUM: acl: fix regression introduced by latest converters support
- MINOR: connection: clear errno prior to checking for errors
- BUG/MINOR: checks: do not trust errno in write event before any syscall
- MEDIUM: checks: centralize error reporting
- OPTIM: checks: don't poll on recv when using plain TCP connects
- OPTIM: checks: avoid setting SO_LINGER twice
- MINOR: tools: add a generic binary hex string parser
- BUG/MEDIUM: checks: tcp-check: do not poll when there's nothing to send
- BUG/MEDIUM: check: tcp-check might miss some outgoing data when socket buffers are full
- BUG/MEDIUM: args: fix double free on error path in argument expression parser
- BUG/MINOR: acl: fix sample expression error reporting
- BUG/MINOR: checks: tcp-check actions are enums, not flags
- MEDIUM: checks: make tcp-check perform multiple send() at once
- BUG/MEDIUM: stick: completely remove the unused flag from the store entries
- OPTIM: ebtree: pack the struct eb_node to avoid holes on 64-bit
- BUG/MEDIUM: stick-tables: complete the latest fix about store-responses
- CLEANUP: stream_interface: remove unused field err_loc
- MEDIUM: stats: don't use conn->xprt_st anymore
- MINOR: session: add a simple function to retrieve a session from a task
- MEDIUM: stats: don't use conn->xprt_ctx anymore
- MEDIUM: peers: don't rely on conn->xprt_ctx anymore
- MINOR: http: prevent smp_fetch_url_{ip,port} from using si->conn
- MINOR: connection: make it easier to emit proxy protocol for unknown addresses
- MEDIUM: stats: prepare the HTTP stats I/O handler to support more states
- MAJOR: stats: move the HTTP stats handling to its applet
- MEDIUM: stats: move request argument processing to the final step
- MEDIUM: session: detect applets from the session by using s->target
- MAJOR: session: check for a connection to an applet in sess_prepare_conn_req()
- MAJOR: session: pass applet return traffic through the response analysers
- MEDIUM: stream-int: split the shutr/shutw functions between applet and conn
- MINOR: stream-int: make the shutr/shutw functions void
- MINOR: obj: provide a safe and an unsafe access to pointed objects
- MINOR: connection: add a field to store an object type
- MINOR: connection: always initialize conn->objt_type to OBJ_TYPE_CONN
- MEDIUM: stream interface: move the peers' ptr into the applet context
- MINOR: stream-interface: move the applet context to its own struct
- MINOR: obj: introduce a new type appctx
- MINOR: stream-int: rename ->applet to ->appctx
- MINOR: stream-int: split si_prepare_embedded into si_prepare_none and si_prepare_applet
- MINOR: stream-int: add a new pointer to the end point
- MEDIUM: stream-interface: set the pointer to the applet into the applet context
- MAJOR: stream interface: remove the ->release function pointer
- MEDIUM: stream-int: make ->end point to the connection or the appctx
- CLEANUP: stream-int: remove obsolete si_ctrl function
- MAJOR: stream-int: stop using si->conn and use si->end instead
- MEDIUM: stream-int: do not allocate a connection in parallel to applets
- MEDIUM: session: attach incoming connection to target on embryonic sessions
- MINOR: connection: add conn_init() to (re)initialize a connection
- MINOR: checks: call conn_init() to properly initialize the connection.
- MINOR: peers: make use of conn_init() to initialize the connection
- MINOR: session: use conn_init() to initialize the connections
- MINOR: http: use conn_init() to reinitialize the server connection
- MEDIUM: connection: replace conn_prepare with conn_assign
- MINOR: get rid of si_takeover_conn()
- MINOR: connection: add conn_new() / conn_free()
- MAJOR: connection: add two new flags to indicate readiness of control/transport
- MINOR: stream-interface: introduce si_reset() and si_set_state()
- MINOR: connection: reintroduce conn_prepare to set the protocol and transport
- MINOR: connection: replace conn_assign with conn_attach
- MEDIUM: stream-interface: introduce si_attach_conn to replace si_prepare_conn
- MAJOR: stream interface: dynamically allocate the outgoing connection
- MEDIUM: connection: move the send_proxy offset to the connection
- MINOR: connection: check for send_proxy during the connect(), not the SI
- MEDIUM: connection: merge the send_proxy and local_send_proxy calls
- MEDIUM: stream-int: replace occurrences of si->appctx with si_appctx()
- MEDIUM: stream-int: return the allocated appctx in stream_int_register_handler()
- MAJOR: stream-interface: dynamically allocate the applet context
- MEDIUM: session: automatically register the applet designated by the target
- MEDIUM: stats: delay appctx initialization
- CLEANUP: peers: use less confusing state/status code names
- MEDIUM: peers: delay appctx initialization
- MINOR: stats: provide some appctx information in "show sess all"
- DIET/MINOR: obj: pack the obj_type enum to 8 bits
- DIET/MINOR: connection: rearrange a few fields to save 8 bytes in the struct
- DIET/MINOR: listener: rearrange a few fields in struct listener to save 16 bytes
- DIET/MINOR: proxy: rearrange a few fields in struct proxy to save 16 bytes
- DIET/MINOR: session: reduce the struct session size by 8 bytes
- DIET/MINOR: stream-int: rearrange a few fields in struct stream_interface to save 8 bytes
- DIET/MINOR: http: reduce the size of struct http_txn by 8 bytes
- MINOR: http: switch the http state to an enum
- MINOR: http: use an enum for the auth method in http_auth_data
- DIET/MINOR: task: reduce struct task size by 8 bytes
- MINOR: stream_interface: add reporting of ressouce allocation errors
- MINOR: session: report lack of resources using the new stream-interface's error code
- BUILD: simplify the date and version retrieval in the makefile
- BUILD: prepare the makefile to skip format lines in SUBVERS and VERDATE
- BUILD: use format tags in VERDATE and SUBVERS files
- BUG/MEDIUM: channel: bo_getline() must wait for \n until buffer is full
- CLEANUP: check: server port is unsigned
- BUG/MEDIUM: checks: agent doesn't get the response if server does not closes
- MINOR: tools: buf2ip6 must not modify output on failure
- MINOR: pattern: do not assign SMP_TYPES by default to patterns
- MINOR: sample: make sample_parse_expr() use memprintf() to report parse errors
- MINOR: arg: improve wording on error reporting
- BUG/MEDIUM: sample: simplify and fix the argument parsing
- MEDIUM: acl: fix the argument parser to let the lower layer report detailed errors
- MEDIUM: acl: fix the initialization order of the ACL expression
- CLEANUP: acl: remove useless blind copy-paste from sample converters
- TESTS: add regression tests for ACL and sample expression parsers
- BUILD: time: adapt the type of TV_ETERNITY to the local system
- MINOR: chunks: allocate the trash chunks before parsing the config
- BUILD: definitely silence some stupid GCC warnings
- MINOR: chunks: always initialize the output chunk in get_trash_chunk()
- MINOR: checks: improve handling of the servers tracking chain
- REORG: checks: retrieve the check-specific defines from server.h to checks.h
- MINOR: checks: use an enum instead of flags to report a check result
- MINOR: checks: rename the state flags
- MINOR: checks: replace state DISABLED with CONFIGURED and ENABLED
- MINOR: checks: use check->state instead of srv->state & SRV_CHECKED
- MINOR: checks: fix agent check interval computation
- MINOR: checks: add a PAUSED state for the checks
- MINOR: checks: create the agent tasks even when no check is configured
- MINOR: checks: add a flag to indicate what check is an agent
- MEDIUM: checks: enable agent checks even if health checks are disabled
- BUG/MEDIUM: checks: ensure we can enable a server after boot
- BUG/MEDIUM: checks: tracking servers must not inherit the MAINT flag
- BUG/MAJOR: session: repair tcp-request connection rules
- BUILD: fix SUBVERS extraction in the Makefile
- BUILD: pattern: silence a warning about uninitialized value
- BUILD: log: fix build warning on Solaris
- BUILD: dumpstats: fix build error on Solaris
- DOC: move option pgsql-check to the correct place
- DOC: move option tcp-check to the proper place
- MINOR: connection: add simple functions to report connection readiness
- MEDIUM: connection: centralize handling of nolinger in fd management
- OPTIM: http: set CF_READ_DONTWAIT on response message
- OPTIM: http: do not re-enable reading on client side while closing the server side
- MINOR: config: add option http-keep-alive
- MEDIUM: connection: inform si_alloc_conn() whether existing conn is OK or not
- MAJOR: stream-int: handle the connection reuse in si_connect()
- MAJOR: http: add the keep-alive transition on the server side
- MAJOR: backend: enable connection reuse
- MINOR: http: add option prefer-last-server
- MEDIUM: http: do not report connection errors for second and further requests
In HTTP keep-alive, if we face a connection error to the server while sending
the request, the error should not be reported, and the client-side connection
should simply be closed, so that client knows it can retry. This can happen if
the server has too short a keep-alive timeout and quits at the same moment the
new request comes in.
When the load balancing algorithm in use is not deterministic, and a previous
request was sent to a server to which haproxy still holds a connection, it is
sometimes desirable that subsequent requests on a same session go to the same
server as much as possible. Note that this is different from persistence, as
we only indicate a preference which haproxy tries to apply without any form
of warranty. The real use is for keep-alive connections sent to servers. When
this option is used, haproxy will try to reuse the same connection that is
attached to the server instead of rebalancing to another server, causing a
close of the connection. This can make sense for static file servers. It does
not make much sense to use this in combination with hashing algorithms.
This commit allows an existing server-side connection to be reused if
it matches the same target. Basic controls are performed ; right now
we do not allow to reuse a connection when dynamic source binding is
in use or when the destination address or port is dynamic (eg: proxy
mode). Later we'll have to also disable connection sharing when PROXY
protocol is being used or when non-idempotent requests are processed.
When a connection to the server is complete, if the transaction
requests keep-alive mode, we don't shut the connection and we just
reinitialize the stream interface in order to be able to reuse the
connection afterwards.
Note that the server connection count is decremented, just like the
backend's, and that we still try to wake up waiters. But that makes
sense considering that we'll eventually be able to immediately pass
idle connections to waiters.
