With this patch, it is possible to configure HAProxy to forge the SSL
certificate sent to a client using the SNI servername. We do it in the SNI
callback.
To enable this feature, you must pass following BIND options:
* ca-sign-file <FILE> : This is the PEM file containing the CA certitifacte and
the CA private key to create and sign server's certificates.
* (optionally) ca-sign-pass <PASS>: This is the CA private key passphrase, if
any.
* generate-certificates: Enable the dynamic generation of certificates for a
listener.
Because generating certificates is expensive, there is a LRU cache to store
them. Its size can be customized by setting the global parameter
'tune.ssl.ssl-ctx-cache-size'.
This patch adds the ssl-dh-param-file global setting. It sets the
default DH parameters that will be used during the SSL/TLS handshake when
ephemeral Diffie-Hellman (DHE) key exchange is used, for all "bind" lines
which do not explicitely define theirs.
This patch does'nt add any new feature: the functional behavior
is the same than version 1.0.
Technical differences:
In this version all updates on different stick tables are
multiplexed on the same tcp session. There is only one established
tcp session per peer whereas in first version there was one established
tcp session per peer and per stick table.
Messages format was reviewed to be more evolutive and to support
further types of data exchange such as SSL sessions or other sticktable's
data types (currently only the sticktable's server id is supported).
Most of the keywords in the global section does not check the maximum
number of arguments. This leds sometines to unused and wrong arguments
in the configuration file. This patch add a maximum argument test in
many keywords of this section.
This patch checks the number of arguments of the keywords:
'global', 'defaults', 'listen', 'backend', 'frontend', 'peers' and
'userlist'
The 'global' section does not take any arguments.
Proxy sections does not support bind address as argument anymore. Those
sections supports only an <id> argument.
The 'defaults' section didn't had any check on its arguments. It takes
an optional <name> argument.
'peers' section takes a <peersect> argument.
'userlist' section takes a <listname> argument.
If the 'userlist' keyword parsing returns an error and no userlist were
previously created. The parsing of 'user' and 'group' leads to NULL
derefence.
The userlist pointer is now tested to prevent this issue.
In order to support http-response redirect, the parsing needs to be
adapted a little bit to only support the "location" type, and to
adjust the log-format parser so that it knows the direction of the
sample fetch calls.
These ones were already obsoleted in 1.4, marked for removal in 1.5,
and not documented anymore. They used to emit warnings, and do still
require quite some code to stay in place. Let's remove them now.
We don't use findproxy_mode() anymore so we can check the conflicting
modes and report the anomalies accordingly with line numbers and more
explicit details.
First, findproxy() was renamed proxy_find_by_name() so that its explicit
that a name is required for the lookup. Second, we give this function
the ability to search for tables if needed. Third we now provide inline
wrappers to pass the appropriate PR_CAP_* flags and to explicitly look
up a frontend, backend or table.
A nasty situation happens when two tables have the same name. Since it
is possible to declare a table in a frontend and another one in a backend,
this situation may happen and result in a random behaviour each time a
table is designated in a "stick" or "track" rule. Let's make sure this
is properly detected and stopped. Such a config will now report :
[ALERT] 145/104933 (31571) : parsing [prx.cfg:36] : stick-table name 't' conflicts with table declared in frontend 't' at prx.cfg:30.
[ALERT] 145/104933 (31571) : Error(s) found in configuration file : prx.cfg
[ALERT] 145/104933 (31571) : Fatal errors found in configuration.
Since 1.4 we used to emit a warning when two frontends or two backends
had the same name. In 1.5 we added the same warning for two peers sections.
In 1.6 we added the same warning for two mailers sections. It's about time
to reject such invalid configurations, the impact they have on the code
complexity is huge and it is becoming a real obstacle to some improvements
such as restoring servers check status across reloads.
Now these errors are reported as fatal errors and will need to be fixed.
Anyway, till now there was no guarantee that what was written was working
as expected since the behaviour is not defined (eg: use_backend with a
name used by two backends leads to undefined behaviour).
Example of output :
[ALERT] 145/104759 (31564) : Parsing [prx.cfg:12]: mailers section 'm' has the same name as another mailers section declared at prx.cfg:10.
[ALERT] 145/104759 (31564) : Parsing [prx.cfg:16]: peers section 'p' has the same name as another peers section declared at prx.cfg:14.
[ALERT] 145/104759 (31564) : Parsing [prx.cfg:21]: frontend 'f' has the same name as another frontend declared at prx.cfg:18.
[ALERT] 145/104759 (31564) : Parsing [prx.cfg:27]: backend 'b' has the same name as another backend declared at prx.cfg:24.
[ALERT] 145/104759 (31564) : Error(s) found in configuration file : prx.cfg
[ALERT] 145/104759 (31564) : Fatal errors found in configuration.
For backend load balancing it sometimes makes sense to redispatch rather
than retrying against the same server. For example, when machines or routers
fail you may not want to waste time retrying against a dead server and
would instead prefer to immediately redispatch against other servers.
This patch allows backend sections to specify that they want to
redispatch on a particular interval. If the interval N is positive the
redispatch occurs on every Nth retry, and if the interval N is negative then
the redispatch occurs on the Nth retry prior to the last retry (-1 is the
default and maintains backwards compatibility). In low latency environments
tuning this setting can save a few hundred milliseconds when backends fail.
Within the listener struct we need to use a reference to the TLS
ticket keys which binds the actual keys with the filename. This will
make it possible to update the keys through the socket
Signed-off-by: Nenad Merdanovic <nmerdan@anine.io>
The method used to skip to next rule in the list is wrong, it assumes
that the list element starts at the same offset as the rule. It happens
to be true on most architectures since the list is the first element for
now but it's definitely wrong. Now the code doesn't crash anymore when
the struct list is moved anywhere else in the struct tcpcheck_rule.
This fix must be backported to 1.5.
This is the most important fix of this series. There's a risk of endless
loop and crashes caused by the fact that we go past the head of the list
when skipping to next rule, without checking if it's still a valid element.
Most of the time, the ->action field is checked, which points to the proxy's
check_req pointer (generally NULL), meaning the element is confused with a
TCPCHK_ACT_SEND action.
The situation was accidently made worse with the addition of tcp-check
comment since it also skips list elements. However, since the action that
makes it go forward is TCPCHK_ACT_COMMENT (3), there's little chance to
see this as a valid pointer, except on 64-bit machines where it can match
the end of a check_req string pointer.
This fix heavily depends on previous cleanup and both must be backported
to 1.5 where the bug is present.
Environment variables were expandables only in adresses.
Now there are expandables everywhere in the configuration file within
double quotes.
This patch breaks compatibility with the previous behavior of
environment variables in adresses, you must enclose adresses with double
quotes to make it work.
tcpcheck error messages include the step id where the error occurs.
In some cases, this is not enough. Now, HAProxy also use the comment
field of the latest tcpcheck rule which has been run.
This commit allows HAProxy to parse a new directive in the tcpcheck
ruleset: 'comment'.
It is used to setup comments on the current tcpcheck rules.
This patch introduces quoting which allows to write configuration string
including spaces without escaping them.
Strong (with single quotes) and weak (with double quotes) quoting are
supported. Weak quoting supports escaping and special characters when
strong quoting does not interpret anything.
This patch could break configuration files where ' and " where used.
Chad Lavoie reported an interesting regression caused by the latest
updates to automatically detect the processes a peers section runs on.
It turns out that if a config has neither nbproc nor a bind-process
statement and depending on the frontend->backend chaining, it is possible
to evade all bind_proc propagations, resulting in assigning only ~0UL (all
processes, which is 32 or 64) without ever restricting it to nbproc. It
was not visible in backends until they started to reference peers sections
which saw themselves with 64 processes at once.
This patch addresses this by replacing all those ~0UL with nbits(nbproc).
That way all "bind-process" settings *default* to the number of processes
defined in nbproc instead of 32 or 64.
This fix could possibly be backported into 1.5, though there is no indication
that this bug could have any effect there.
It is sometimes desirable to wait for the body of an HTTP request before
taking a decision. This is what is being done by "balance url_param" for
example. The first use case is to buffer requests from slow clients before
connecting to the server. Another use case consists in taking the routing
decision based on the request body's contents. This option placed in a
frontend or backend forces the HTTP processing to wait until either the whole
body is received, or the request buffer is full, or the first chunk is
complete in case of chunked encoding. It can have undesired side effects with
some applications abusing HTTP by expecting unbufferred transmissions between
the frontend and the backend, so this should definitely not be used by
default.
Note that it would not work for the response because we don't reset the
message state before starting to forward. For the response we need to
1) reset the message state to MSG_100_SENT or BODY , and 2) to reset
body_len in case of chunked encoding to avoid counting it twice.
If a peers section is bound to no process, it's silently discarded. If its
bound to multiple processes, an error is emitted and the process will not
start.
Sometimes it's very hard to disable the use of peers because an empty
section is not valid, so it is necessary to comment out all references
to the section, and not to forget to restore them in the same state
after the operation.
Let's add a "disabled" keyword just like for proxies. A ->state member
in the peers struct is even present for this purpose but was never used
at all.
Maybe it would make sense to backport this to 1.5 as it's really cumbersome
there.
It's dangerous to initialize stick-tables before peers because they
start a task that cannot be stopped before we know if the peers need
to be disabled and destroyed. Move this after.
If a table in a disabled proxy references a peers section, the peers
name is not resolved to a pointer to a table, but since it belongs to
a union, it can later be dereferenced. Right now it seems it cannot
happen, but it definitely will after the pending changes.
It doesn't cost anything to backport this into 1.5, it will make gdb
sessions less head-scratching.
Recently some browsers started to implement a "pre-connect" feature
consisting in speculatively connecting to some recently visited web sites
just in case the user would like to visit them. This results in many
connections being established to web sites, which end up in 408 Request
Timeout if the timeout strikes first, or 400 Bad Request when the browser
decides to close them first. These ones pollute the log and feed the error
counters. There was already "option dontlognull" but it's insufficient in
this case. Instead, this option does the following things :
- prevent any 400/408 message from being sent to the client if nothing
was received over a connection before it was closed ;
- prevent any log from being emitted in this situation ;
- prevent any error counter from being incremented
That way the empty connection is silently ignored. Note that it is better
not to use this unless it is clear that it is needed, because it will hide
real problems. The most common reason for not receiving a request and seeing
a 408 is due to an MTU inconsistency between the client and an intermediary
element such as a VPN, which blocks too large packets. These issues are
generally seen with POST requests as well as GET with large cookies. The logs
are often the only way to detect them.
This patch should be backported to 1.5 since it avoids false alerts and
makes it easier to monitor haproxy's status.
The principle of this cache is to have a global cache for all pattern
matching operations which rely on lists (reg, sub, dir, dom, ...). The
input data, the expression and a random seed are used as a hashing key.
The cached entries contains a pointer to the expression and a revision
number for that expression so that we don't accidently used obsolete
data after a pattern update or a very unlikely hash collision.
Regarding the risk of collisions, 10k entries at 10k req/s mean 1% risk
of a collision after 60 years, that's already much less than the memory's
reliability in most machines and more durable than most admin's life
expectancy. A collision will result in a valid result to be returned
for a different entry from the same list. If this is not acceptable,
the cache can be disabled using tune.pattern.cache-size.
A test on a file containing 10k small regex showed that the regex
matching was limited to 6k/s instead of 70k with regular strings.
When enabling the LRU cache, the performance was back to 70k/s.
This concerns everythins related to accepting a new session and
expiring the embryonic session. There's still a hard-coded call
to stream_accept_session() which could be set somewhere in the
frontend, but for now it's not a problem.
With HTTP/2, we'll have to support multiplexed streams. A stream is in
fact the largest part of what we currently call a session, it has buffers,
logs, etc.
In order to catch any error, this commit removes any reference to the
struct session and tries to rename most "session" occurrences in function
names to "stream" and "sess" to "strm" when that's related to a session.
The files stream.{c,h} were added and session.{c,h} removed.
The session will be reintroduced later and a few parts of the stream
will progressively be moved overthere. It will more or less contain
only what we need in an embryonic session.
Sample fetch functions and converters will have to change a bit so
that they'll use an L5 (session) instead of what's currently called
"L4" which is in fact L6 for now.
Once all changes are completed, we should see approximately this :
L7 - http_txn
L6 - stream
L5 - session
L4 - connection | applet
There will be at most one http_txn per stream, and a same session will
possibly be referenced by multiple streams. A connection will point to
a session and to a stream. The session will hold all the information
we need to keep even when we don't yet have a stream.
Some more cleanup is needed because some code was already far from
being clean. The server queue management still refers to sessions at
many places while comments talk about connections. This will have to
be cleaned up once we have a server-side connection pool manager.
Stream flags "SN_*" still need to be renamed, it doesn't seem like
any of them will need to move to the session.
This will be useful later to state that some listeners have to use
certain decoders (typically an HTTP/2 decoder) regardless of the
regular processing applied to other listeners. For now it simply
defaults to the frontend's default target, and it is used by the
session.
Listerner->timeout is a vestigal thing going back to 2007 or so. It
used to only be used by stats and peers frontends to hold a pointer
to the proxy's client timeout. Now that we use regular frontends, we
don't use it anymore.
The peers frontend timeout was mistakenly set on timeout.connect instead
of timeout.client, resulting in no timeout being applied to the peers
connections. The impact is just that peers can establish connections and
remain connected until they speak. Once they start speaking, only one of
them will still be accepted, and old sessions will be killed, so the
problem is limited. This fix should however be backported to 1.5 since
it was introduced in 1.5-dev3 with peers.
Until now, the TLS ticket keys couldn't have been configured and
shared between multiple instances or multiple servers running HAproxy.
The result was that if a request got a TLS ticket from one instance/server
and it hits another one afterwards, it will have to go through the full
SSL handshake and negotation.
This patch enables adding a ticket file to the bind line, which will be
used for all SSL contexts created from that bind line. We can use the
same file on all instances or servers to mitigate this issue and have
consistent TLS tickets assigned. Clients will no longer have to negotiate
every time they change the handling process.
Signed-off-by: Nenad Merdanovic <nmerdan@anine.io>
This patch adds a new option which allows configuration of the maximum
log level of messages for which email alerts will be sent.
The default is alert which is more restrictive than
the current code which sends email alerts for all priorities.
That behaviour may be configured using the new configuration
option to set the maximum level to notice or greater.
email-alert level notice
Signed-off-by: Simon Horman <horms@verge.net.au>
