Commit Graph

3243 Commits

Author SHA1 Message Date
c26b5a3dd0 Tidy earlier fix and apply another - both follow the principle of not removing the link from session to client DCB when dcb_close is called for client DCB. The close will not fully complete, but will be finalised when the session is freed. At this time, no DCBs will be referring to the session. These measures protect the dcb->data in the client DCB from being lost until the session is no longer needed. 2016-02-24 15:34:02 +00:00
5122777829 Try to fix problem if balancing free client DCB and free session so that auth data is always available and client DCB is not freed until session is ready to be freed. Also fix problem in auth logic. 2016-02-24 10:00:45 +00:00
cb2e3b898e Cleaned up qlafilter, topfilter and namedserverfilter instance creation
The change fixes a crash when no filebase parameter is given for the filter.

The instance creation functions of these filters had spossible memory leaks in
case errors in the configuration occurred. In addition to these, they would be
successfully created even if unexpected parameters were given.
2016-02-23 21:45:31 +02:00
aec1310c52 MXS-580: Added more options for filters
The tee, qlafilter, namedserverfilter and topfilter now support the same filter
options: case, ignorecase and extended. The extended filter option enables
the Extended Regular Expression syntax for the filters which is used when
matching queries.
2016-02-23 20:52:53 +02:00
19356be848 Updated documentation about Bison and Flex dependency
The dbfwfilter now requires Bison and Flex to be built
2016-02-23 20:52:52 +02:00
395cbdc103 Simplify handling of "client data" by having a pointer to it only in DCB and not in session structure. Change name of session->client to session->client_dcb for greater clarity. Temporary free of client data in DCB, to be moved to authenticator when it becomes a module. Fix incorrect name of listener_alloc. 2016-02-23 11:13:21 +00:00
1adbe3b23a No user definitions in dbfwfilter rules is treated as an error again
The lack of user definition lines in the dbfwfilter rule file was ignored
previously but this change makes the lack of users an error again. Also
added function documentation to some of the functions.
2016-02-23 10:34:12 +02:00
ef8a20cceb Fix silly mistakes. 2016-02-22 11:16:51 +00:00
5077933e41 Fix bug in mysql_client.c (over optimisation of protocol setting); various clarifications and improvements re code review. 2016-02-22 11:05:02 +00:00
28bd7c2202 Merge branch '1.2.1-binlog_router_trx' into develop 2016-02-19 18:51:59 +02:00
a947b33769 MXS-585: Fixed buffer length instead of packet length being used in a calculation
The usage of the buffer length in the authentication success checks causes
the authentication to fail with routers that don't require statement input.
2016-02-18 18:03:21 +02:00
c25ef2f72a Added missing 5.5.5- prefix to MaxScale MySQL version
The handshake was sending "MaxScale 1.3.0" as the server version which causes
some connectors to think that the version is not valid.
2016-02-18 18:03:21 +02:00
effa2dbdae Set dbfwfilter maturity to GA and set the version to 1.1.0
The module has been GA for a while and it has also changed so the version
should be updated.
2016-02-17 10:59:26 +02:00
497a72d42f Added extra logging to dbfwfilter output
The real username and host is now logged in addition to the user definition
that it matched. The messages also have the service name to distinct
connections to two different services which both use the same filter.
2016-02-17 10:45:21 +02:00
59f5880898 Added missing OK byte to payload size calculation
The OK byte was not taken into notice when the total size of all the payloads
in all the packets was calculated.
2016-02-17 08:30:50 +02:00
63ce9fe6bc Fixed formatting and added more error checks
Added log messages when ftruncate fails and cleaned up formatting.
2016-02-16 13:06:25 +02:00
cd2af6ffef Cleaned up the code based on the code review
Added missing error condition checks and cleaned up code.
2016-02-16 13:06:25 +02:00
a55f017c75 Fixed packets with a length of one being ignored
The packets were not written into the binlogs which caused binlog corruption.
2016-02-16 13:06:25 +02:00
9306b9d68c Added detection of checksums split across two packets
The checksums should now be processed properly event if the event is in more than
one packet.
2016-02-16 13:06:25 +02:00
36896afcbd Fixed missing NULL check when reading records
If the binlog record was not found, a NULL pointer is returned. There was no
check for the return value and it assumed that it was always non-NULL.
2016-02-16 13:06:25 +02:00
74c8b5e296 Fixed events larger than 2^24 failing without transaction safety
If transaction safety was disabled and a large event sent in multiple SQL
packets was received, the distribution of that event to the slaves would fail.
2016-02-16 13:06:25 +02:00
2b7e2d3043 Added checksum calculations for events larger than 2^24 bytes
The checksums are now properly calculated for large events that span multiple
SQL packets.
2016-02-16 13:06:25 +02:00
3609f97ba0 Fixed events which are exactly 0x00fffffe bytes long failing to replicate
The empty packet sent after a large event which fits into exactly one packet
was written to disk and the writing of no bytes caused it to be treated as
an error.
2016-02-16 13:06:25 +02:00
476691eda1 Removed log message for event larger than 16MB
The log message used during tests is now removed
2016-02-16 13:06:25 +02:00
3e04a36ac3 Added support for distribution of packets larger than 2^24 bytes
Moved the the sending of the replication events to a different function
and added support for events that span multiple MySQL packets.
2016-02-16 13:06:25 +02:00
12ee568978 Fixed last_written being set to the size of the event
The addition used =+ instead of += which caused it to be an assignment.
2016-02-16 13:06:25 +02:00
d2b4713d27 Added missing condition to else clause
This fixes all packets being considered as large packets.
2016-02-16 13:06:25 +02:00
ae33df3cbc Large events are now processed in chuncks
The router->last_written is used to store the position where the last event was
written. The replication header is also stored in a separate structure in
the router which is used later when the last packet of a multi-packet event
arrives.
2016-02-16 13:06:24 +02:00
d3e1d4dd2f First fix for 16MB handling in the master part
First fix for 16MB handling in the master part.

Distribute events to up to date slave is not included yet
2016-02-16 13:06:24 +02:00
ab1fb90d86 Fix for 16MB transmission in the slave part
Fix for 16MB transmission in the slave part
2016-02-16 13:06:24 +02:00
866e91c088 Changes to improve robustness of SSL processing, separate it so far as possible from protocols. Separate the detailed mechanism of authentication from the MySQL protocol with a view to making it into a module later. 2016-02-15 09:29:07 +00:00
26cbcb7365 MXS-511: Made log messages about master status coherent
The checking of the master status and the possible error logging were done
in two different steps. This led to confusing error messages when the state
of the server changed between the check and the logging of the error message.
2016-02-12 21:18:57 +02:00
e777828bd0 When creating heartbeat packet too many bytes were copied.
The memory area ‘ptr’ points to contains now the right data
2016-02-11 17:12:10 +01:00
4eccc1acb6 When creating heartbeat packet too many bytes were copied.
The memory area ‘ptr’ points to contains now the right data
2016-02-11 17:06:13 +01:00
cfefc046e9 Merge branch 'release-1.3.0' into develop 2016-02-11 10:23:00 +02:00
eaddc4ea9d Fixed dbfwfilter build failure on Ubuntu
The bison executable didn't generate the headers even though it was instructed
to generate them. This was caused by an older CMake version.
2016-02-09 16:35:51 +02:00
9e33f300b0 Added flex/bison based parsing of rules to dbfwfilter
The dbfwfilter now uses flex to tokenize the rules and bison to process those
tokens into a language. This will make the syntax checks more robust and
the rule language easier to maintain. It also makes the processing of rules
easier to understand and removes most of the C language side checks for
correctness.

The old rule processing functions are still left intact and will be removed
at a later point.
2016-02-09 11:00:27 +02:00
6055ba25bf Added logging of matching and non-matching queries to dbfwfilter
When configured to log matching queries, the dbfwfilter will log all queries
that match a rule. The rule name, user name and the query itself are logged.

It is also possible to match all queries that do not match a rule. Only the user
name and query is logged in this mode.
2016-02-09 11:00:27 +02:00
5efd5d8927 Added user configurable actions to dbfwfilter
The dbfwfilter can now either allow, block or ignore queries that match the
configured rules.
2016-02-09 11:00:27 +02:00
9acdd4e70c Fixed tests failing to build
The tests were not linking against the maxscale-common library.
2016-02-09 10:59:11 +02:00
4a4f22c9a7 Make query_classifier part of core.
The query classifier interface is now built as an integral
part of MaxScale core.
2016-02-08 11:44:07 +02:00
834d04ed6c Merge branch 'MXS-517' into develop 2016-02-08 11:03:31 +02:00
379cb6c781 MXS-564: Slave server states were not properly cleared after session commands
When a command that changes the session state is executed on all servers and
a slave server goes down, an error message was sent to the client even though
one wasn't expected. These appeared as odd connection errors on the client side.
2016-02-05 13:54:24 +02:00
9432967da5 Don't assert the master server state when routing query to slave
When a query is being routed to the slave, it is not necessary to assert that
the master server's state is still master. It is possible that the master server
changes states while the query is being routed but that doesn't affect the query
being currently routed.
2016-02-04 09:57:27 +02:00
e731f6a29e Fixed double freeing of buffers
The buffers in readwritesplit are handled at the routeQuery level and freeing
them at lower levels cause a double free of the used buffers.
2016-02-04 09:57:27 +02:00
d9eeb72198 Added missing check for closed master backend reference into readwritesplit
The master DCB was checked for NULL-ness but the proper way is to check if
the backend reference is closed. This will fix a debug assertion in addition
to possibly preventing crashes.
2016-02-04 09:57:27 +02:00
971d1d5de2 Redesign logic of SSL connections to work with status in client DCB instead of in the protocol. This eases the way to moving SSL logic out of a specific protocol (currently MySQL) so as to be available across any protocol. Also, some simplification. 2016-02-03 08:54:55 +00:00
e7fb2640d3 Fixed authentication failing due to unknown database even though it exists
When the shard maps are being updated they are set into a stale state. This
means that one client connection is updating the shard maps and the information
in the shard map is not the most recent. This does mean that the information
is valid and authentication should succeed even if the shard map is stale.
2016-02-03 00:47:06 +02:00
081bacdc64 Switched to MySQL client library headers for MaxScale
The include directories previously used by MaxScale were from the embedded
library. All parts of MaxScale apart from the query classifier now use
the client libraries.
2016-02-02 14:04:59 +02:00
0d51f78a2a Dbfwfilter now uses PCRE2 regular expressions
Switched from the POSIX regex implementation to PCRE2. This should improve the
filter's performance as well as make the code more manageable.
2016-02-02 12:38:44 +02:00