| Age | Commit message (Collapse) | Author |
|---|
|
This patch adds the --any-nat option that allows to display src-NATted
OR dst-NATted flows.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch is a cleanup, it puts all the commands and options code
together. This makes easier and less error-prone the task to add
new commands and options.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch fixes a hung that occurs if you invoke `conntrackd -c'
and you have disabled the external cache.
Note that `conntrackd -c' does nothing since there is no entries
in the external cache to be committed.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds a new option TCPWindowTracking that allows not
to disable TCP window tracking as it occurs by default.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds state-synchronization for ICMP. You SHOULD use a
Linux kernel >= 2.6.31, otherwise this patch can result in tons
of state-updates.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the clause `DisableInternalCache' that allows you
to bypass the internal cache. This clause can only be used with
the notrack synchronization mode.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
Under stress, the TCP stack may return EAGAIN if there is not
space left in the sender buffer. We also enqueue any other
error.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch fixes a bug in the TCP support that breaks
re-connections of the client side if several TCP
channels are used in the configuration file.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
Use the TCP header size (20 bytes) instead of the UDP header size
(8 bytes) to calculate the maximum packet size.
Reported-by: Samuel Gauthier <samuel.gauthier@6wind.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
We cannot assume that we will not write in the net message before we
send it, because the memory allocated for the net message (__net) is
only reserved in BUILD_NETMSG (because of the { } block in it).
This patch marks the buffer as static to avoid this problem.
Based on a patch from Samuel Gauthier <samuel.gauthier@6wind.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds support for TCP as protocol to replicate
state-changes between two daemons. Note that this only
makes sense with the notrack mode.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
With this patch, we can remove file descriptors dinamically
from our own file descriptor pool.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch reduces the number of gettimeofday syscalls by caching
the current time in a variable at the beginning of the main loop.
Based on a suggestion from Vincent Jardin.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the clause `DisableExternalCache' that allows you
to disable the external cache and to directly inject the entries
into the kernel conntrack table. As a result, the CPU consumption
of conntrackd increases. This clause can only be used with the
FT-FW and the notrack synchronization modes, but not with the
alarm mode.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the missing support to filter IPv6 from kernel-space
by means of the BSF API that libnetfilter_conntrack provides.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the NetlinkEventsReliable clause, this is useful
to turn on reliable Netlink event delivery. This features
requires a Linux kernel >= 2.6.31.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch reworks the commit phase to avoid the forking. This is
particularly useful in active-active setups in which one node
has to commit the external cache while it is receiving new entries
to be added in the external cache. This results in really high
commit times due to the penalty of the copy-on-write that fork
performs.
The default number of steps in one run loop is limited to 64 by now.
No option to tune this parameter is still available via the
configuration file.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds cache_iterate_limit() and hashtable_iterate_limit()
that allows to limit the iteration to given a number of states.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the LOCAL_RET_* return values. The return value
LOCAL_RET_STOLEN which allows to leave a client socket open while
waiting for an operation to finish.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds a new statistic field to count the number of
enospc errors while adding new nodes to some queue.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch re-introduces `-s queue' but now it displays generic
queue statistics.
# conntrackd -s queue
active queue objects: 0
queue txqueue:
current elements: 0
maximum elements: 2147483647
not enough space errors: 0
queue rsqueue:
current elements: 72
maximum elements: 128
not enough space errors: 0
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the name field to identify the queue by means of
a string. This patch is used by the next one that introduces
per-queue statistics.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch renames the statistics option that displays the content
of the resend queue which is used by the ftfw mode.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
In 400ae54438c4b85126f9fab0ae1dc067823b70f7, we added the SCTP
support by means of a structure that was encapsulated in an
TLV attribute. However, this structure didn't handle alignment
and endianess issues appropriately. Similar problem was
introduced in b808645ec71b7cc22cf5106b3d79625d07e6077c along
with the DCCP support.
This patch moves every field of this structure to independent
attributes. I decided not to use nesting to make building and
parsing more simple.
Using TLV is a good idea, specially for DCCP and SCTP that are
under development and that may include new fields and obsolete
them in the future.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds the ability to dump the list of existing child
processes. In general, it would be hard to display one since
child processes are generally forked for very specific tasks,
like commit and flush operations, and they have very limited
lifetime. However, this can be handy for debugging problems.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds a dedicated commit handler since there is a possible
race condition that can happen if the child process ends before we
have received all the event messages that the commit request has
triggered.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
In 6f5666a29cb7cbff08ce926ee1edb84a311ff6ee, I moved the flush
operation into a child process and to use a disposable handler
to perform flush requests. This patch adds a dedicated flush
handler since there is a possible race condition that can
happen if the child process ends before we have received all
the event messages that the flush request has triggered.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch allows to limit the number of simultaneous child processes.
This is required by the next patch that replaces disposable handlers
to commit and flush with permanent handlers.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch is a cleanup, it removes an unused nfct handler. This
removal is due to recent commits that has obsolete it.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch removes a reminiscent of the lifetime cache feature.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch removes the cache write-through clause. This feature
remained undocumented although some has found it looking at the
source code. This feature has remained in the tree for quite
some time although it has several limitations. Moreover, it
is specifically broken and dangerous for Linux kernels >= 2.6.29
since it generates loops in the synchronization.
We do this removal first to prepare the introduction of a feature
to bypass the external cache.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
With this patch, the flush operation is performed by a child process.
Thus, the parent process digests destroy events that ctnetlink reports
back and, thanks to the origin infrastructure, we skip the messy
implicit synchronization that are triggered by such events.
This patch requires a Linux kernel >= 2.6.29 to benefit from this
change, otherwise it has no effect.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
Since Linux kernel 2.6.29, ctnetlink reports the changes that have
been done using ctnetlink. With this patch, conntrackd can recognize
who is the origin of the event messages. For example, this is
interesting to avoid a messy implicit bulk send during the commit
of entries.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds a simple infrastructure that allows to account
the child processes that have been forked. This also includes
a callback handler that can be registered that is called once
the child process finishes.
We can extended this later to include an alarm to limit the
maximum lifetime of a forked child process. This is good to
ensure that child processes behave timely.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds initial support for DCCP state replication.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds initial support for SCTP state replication.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds GRE support for the command line tool conntrack.
With this patch, we support all protocols available in the kernel.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.
|
|
With this patch, you can change the scheduler policy and priority
for conntrackd. Using a RT scheduler policy reduces the chances to
hit ENOBUFS in Netlink.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds DCCP support for the command line tool conntrack.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds SCTP support to the command line tool conntrack.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds UDPlite support for the command line tool conntrack.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch extends the generic_opt_check() function to add
extra information on the possible option combinations. Under
some specific situations, like the creation and getting of
a conntrack, you may specify the original or the reply tuple
but at least one MUST be present. This handling has been
always tricky, it still remains but we're more user friendly
at least.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch removes a couple of constants that have no clients in
the conntrackd code.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch adds `-S' command to display kernel statistics. Using
raw `cat' on /proc and the hexadecimal output is not very handy.
This option parses the /proc entry and display the information
is a more human friendly way.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch fixes a problem that was introduced while adding the
multichannel support.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch obsoletes the `ListenTo' clause which is a reminiscent
of the intial event filtering code.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch removes several structure fields that are unused.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch fixes an EINVAL error returned by bind() when opening
an UDP server socket to propagate state-changes over the dedicated
link. This patch also includes the change of the example
configuration files in case that you want to use UDP over IPv6.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch obsoletes `-s multicast' by `-s link' to display the
dedicated link statistics, as the current dedicated link protocol
use can be unicast UDP or multicast. The term "link" is more
generic.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This patch removes a compilation warning. The buffer passed to
be sent must be const.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
