| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
| |
This patch modifies the SQL schema for MySQL and PostgreSQL to store
the mac address only once (instead of duplicating the mac address for
each packet). This is done by using a shared reference to the entry
containing the tuple (mac_address,mac_protocol).
Signed-off-by: Pierre Chifflier <chifflier@inl.fr>
|
|
|
|
|
|
|
|
| |
This patch fixes some messages in the NFCT and NFLOG input
plugin (end of line before quote). It also fixes indenting by
suppressing some spaces on empty line and replacing spaces by tab.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
Add stack example for MARK and update some wrong stacks.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
|
| |
This patch modifies plugins to use the already defined but not used
define. This also fixes some weird behaviours in error treatment (like
not stopping after OOM).
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
|
| |
The config file parsing was not able to parse integer given in hex notation.
This patch modify the parsing of configfile to be able to use different
integers notation.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
|
|
| |
This module filters message by using the mark to decide wether or not a
packet or a flow has to be logged. It takes a mark and a mask option. It
demonstrates the usage of ULOGD_IRET_STOP which can be used to abort
iteration through the stack.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
|
|
| |
When a plugin returns ULOGD_IRET_STOP, the propagation should
stop. This was not the case as break was used to do so but it was called
inside a switch and thus apply to the switch instruction and not to
the llist iteration.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
| |
crash due to ulogd_logfile set to a string allocated on stack by config_parse_file
|
|
|
|
|
|
|
|
|
| |
This patch fixes NFCT when hash_enable is 0. Limitation of treatment to
NFCT_DESTROY message type causes usage of the hashtable function and
hence a crash because it is not initiated.
Signed-off-by: regit <regit@ghlodit.inl.fr>
Signed-off-by: Patrick McHardy <kaber@trash.net>
|
|
|
|
|
| |
instead of being put in the new type ui128. The result was an improper value
of the IPv6 source add destination addresses.
|
| |
|
|
|
|
| |
is used (hash_enable=1 which is the default)
|
|
|
|
| |
entries depending on the event type parameter.
|
| |
|
|
|
|
| |
STOP timestamp for DESTROY packet
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
This patch improves the overrun handling. The NFLOG plugin duplicates the
netlink buffer size if the size does not goes after the upper boundary.
This patch also introduces two new clauses, the netlink_socket_buffer_size
and netlink_socket_buffer_maxsize that set the size of the netlink socket
buffer.
|
|
|
|
|
|
| |
This patch cleans up the key building by breaking lines at 80 columns and
it fixes the IPv6 support (use of a pointer after free) by introducing a new
128 bit type.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch improves the overrun handling. The logic behind this patch
consists of two steps:
1) duplicate the netlink buffer size if the size does not goes after the
upper boundary.
2) scheduling a resynchronization (in two seconds) with the kernel conntrack
table if we hit ENOBUFS. During the resynchronization, the NFCT plugin dumps
the current table and purges the objects that do not exist anymore.
This patch also introduces two new clauses, the netlink_socket_buffer_size
and netlink_socket_buffer_maxsize that set the size of the netlink socket
buffer.
|
|
|
|
|
|
|
|
| |
This patch introduces a generic hashtable to store the nf_conntrack objects.
The objects are identified by the original and reply tuples instead of the
conntrack ID which is not dumped in the event message of linux kernel < 2.6.25.
This patch also fixes the NFCT_MSG_* by NFCT_T_* which is the appropriate
message type tag.
|
|
|
|
|
|
|
| |
can be used to determine if the packet has been dropped, rejected or accepted.
The meaning of label is completely user-defined.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
| |
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
connections.
Signed-off-by: Pierre Chifflier <chifflier@inl.fr>
|
|
|
|
|
|
| |
Update view_tcp_quad and view_udp_quad.
Signed-off-by: Pierre Chifflier <chifflier@inl.fr>
|
| |
|
|
|
|
| |
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
|
| |
NFLOG has been modified to support GID display. There is a problem as this
feature is only available in latest subversion of libnetfilter_log. This
patch made this feature optional:
* It detects if system support the nflog_get_gid() function
* Compilation of nflog_get_gid() related code is conditional
|
| |
|
| |
|
|
|
|
|
| |
I have no idea what the intention behind this change was, but it
seems bogus, the output format should (mostly) match ipt_LOG.
|
|
|
|
|
|
|
|
|
| |
Due to the modifications done to be able to use multiple time the SOURCE
plugin, a single instance of database output plugin could not anymore be
used in separate stack. This patch fixes this by limiting the effect of
the previous modification on SOURCE plugin.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
| |
Signed-off-by: Anton Vazir <anton.vazir@gmail.com>
|
|
|
|
| |
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
some stack to take my latest patches into account.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
accross the stack NFCT IP2BIN MYSQL.In fact IP2BIN out .bin suffixed
fields.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
This patch introduces a explicit list of input keys and obtains IPv6
compliance by using IP2STR output as input for IP address.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
| |
|
|
|
|
|
|
| |
have now to be used with a defined IP storage type.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
detection.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
plugin. To be able to send IP addresses to the IP2STR and IP2BIN module
oob.family and oob.protocol keys have been added.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
This patch change the input key of the module to use conversion made by the
IP2STR module.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
| |
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
address to be transformed in a string not really related to the real Ipv6
address.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
for flow display.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
misbehaviour was also causing to read datas out of the correct range.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
| |
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
patch also removes mac_daddr which does not provide any interesting
logging information.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
|
| |
of conversion to string of MAC address. It is used by database
output plugin to store MAC related information.
Signed-off-by: Eric Leblond <eric@inl.fr>
|
|
|
|
|
|
| |
interger. This patch fix this in the ULOG module.
Signed-off-by: Eric Leblond <eric@inl.fr>
|