b10fc196a1
Fix maxinfo/maxadmin debug assertion
...
Both services could write to a DCB from a worker that doesn't own it.
2019-05-10 11:23:10 +03:00
603eac9cf8
Fix build failures
...
Narrowing type conversions in schemarouter and avrorouter.
2019-05-10 10:37:52 +03:00
de95fcc9d6
Format rest of the sources
2019-05-10 10:31:12 +03:00
fbd61967a7
Format filters
2019-05-10 10:31:12 +03:00
418ccf861d
Format routers and monitors
2019-05-10 10:31:12 +03:00
6625c1296b
Format authenticator and protocol modules
2019-05-10 10:31:12 +03:00
0d13e694e4
Merge branch '2.3' into develop
2019-05-10 09:28:39 +03:00
b4e8f79c5f
Format core sources again
...
Formatted with nl_func_type_name and related options set to ignore. This
keeps the formatting intact for long return types in declarations and
definitions.
2019-05-10 09:21:52 +03:00
50b5fe76ef
Pass parameters as const ref to server_alloc
2019-05-10 09:21:52 +03:00
d203e7af83
Move SSL related code into ssl.cc
...
The code was in the listener sources but was shared by both listeners and
servers.
2019-05-10 09:21:52 +03:00
31ce757f8d
Don't attempt TLS with maxscaled protocol
...
The authenticator doesn't support it so it is pointess to try it.
2019-05-10 09:21:52 +03:00
3813c728b1
Move listener parameter handling into Listener::create
...
The Listener::create method now takes a set of configuration parameters
from which it constructs a listener. This removes the duplicated code and
makes the behavior of listener creation similar to other objects in
MaxScale. It also allows the configuration parameters to be stored in the
listener object itself.
2019-05-10 09:21:52 +03:00
ca21d27c68
Merge branch 'develop' of github.com:mariadb-corporation/MaxScale into develop
2019-05-10 09:18:32 +03:00
142038051c
Fix GCC 9 warnings
...
GCC 9 has warnings about pessimistic return values which are triggered by
returning a local value via std::move.
2019-05-10 09:11:40 +03:00
1efe3e0b20
Fix avrorouter client notifications
...
The notifications that tell clients about new data weren't using the
correct mechanism.
2019-05-10 09:11:39 +03:00
37bfe1cd67
Never exceed persistent connection limits
...
By doing the persistent connection count incrementation with CAS, value
can be guaranteed to never exceed the configured limit.
2019-05-10 09:11:39 +03:00
034dc6a67e
MXS-2472 Update BLR docs on secondary masters
2019-05-09 17:14:57 +03:00
3c5a932010
Fix masking_mysqltest for 2.3
...
Since UNIONs that would allow a column to be masked are
rejected, there is no need to check that masking is performed
since we will never get that far.
2019-05-09 17:12:52 +03:00
23a09a6294
MXS-2455 Use mxb::Buffer::iterator
...
Simplifies the code and as extra allocations etc. are only
made when info is enabled, and can thus be ignored.
2019-05-09 15:04:03 +03:00
381940ce8c
Merge branch '2.3' into develop
2019-05-09 14:35:44 +03:00
c72dbb6f46
MXS-2448: Fix connection counts of persistent DCBs
...
DCBs that were closed while they were persisted were double decremented.
2019-05-09 13:53:32 +03:00
d2e49c9591
MXS-2448: Assert that connection counts never go negative
...
The connections should never go below zero. It seems that at least pers_01
test suffers from this.
2019-05-09 13:06:27 +03:00
567ad9b8b8
Fix galeramon regression
...
The comparisons were wrong: strcasecmp returns 0 for equal strings.
2019-05-09 12:30:29 +03:00
59f2145c00
Allocate blr heartbeat buffer on correct worker
...
The buffer was allocated on one worker and written on another.
2019-05-09 12:30:29 +03:00
788dc429f8
Do client callback on owning worker
...
The callback should've been done on the worker that owns the DCB instead
of the main worker.
2019-05-09 12:30:26 +03:00
b313c6d0e7
MXS-2474 Ignore attempts to re-register a housekeeper task
...
It is an error to register the same task multiple times, but
for a maintenance release it is simpler and less risky to simply
ignore an attempt (that BLR does) to do that.
Allowing a task to be registered anew causes behaviour akin
to a leak.
2019-05-09 10:58:35 +03:00
0423c5528e
Merge branch '2.3' into develop
2019-05-08 15:56:35 +03:00
5ea2bab88c
Merge branch '2.2' into 2.3
2019-05-08 15:56:08 +03:00
ff1959fda7
Merge branch '2.2.21' into 2.2
2019-05-08 15:55:42 +03:00
73f8bf1f6f
Update 2.2.21 release date
2019-05-08 15:49:11 +03:00
1df6495d9a
Update release notes on MariaDB-Monitor change
2019-05-08 15:32:08 +03:00
e3b5ba9620
MXS-1973 Support reverse DNS for client hostnames in MaxCtrl
...
May slow maxscale down when used. Only supported for "list sessions",
"show sessions" and "show session <id>".
2019-05-08 15:04:44 +03:00
0e0342e657
Merge branch '2.3' into develop
2019-05-08 10:38:39 +03:00
a652b6bd5b
Add advance(int) to mxs::Buffer iterators
...
This makes iterating over packets in buffers faster while still
maintaining the requirements for forward iterators. Not using operator+=
makes it clear that this is not a random access iterator.
2019-05-08 10:33:27 +03:00
bad96e411a
Merge branch '2.3' into develop
2019-05-08 10:30:58 +03:00
c818b1208a
MXS-2455 Recognize transaction rollbacks
...
All transaction rollback errors have an sql_state like "40XXX".
So, when an error reply is received we check for that and act
accordingly.
2019-05-08 10:00:50 +03:00
0638ea736e
Write slave heartbeat in correct thread
...
The writing should be done on the worker that owns the DCB.
2019-05-08 09:44:55 +03:00
446788f2ed
MXS-1799 Add timestamps to retain_last_statements messages
2019-05-07 22:54:31 +03:00
3d66e68e95
MXS-2170 Start MaxScale normally if it gets the same PID as previous
...
MaxScale
Check is made to see if the found MaxScale PID is owned by the process
itself.
2019-05-07 22:23:46 +03:00
56c27afd0a
MXS-2458: add template for Clustrix monitor tests
...
First implementation of Clustrix setup creation
TODO: move Clustrix installation to MDBCI
2019-05-07 17:17:57 +03:00
4e6ffc0381
Clean up server config parameter handling
...
Removes helper classes which are no longer required.
2019-05-07 15:39:34 +03:00
9ae25e67a8
Update maintenance number in 2.2 minor branch
2019-05-07 15:27:48 +03:00
a7a2215681
Update release notes and change log
2019-05-07 15:22:37 +03:00
d3a1f5c51f
MXS-2463 Prefer range loops
...
Also consistently use 'kv' as in key-value as the element name
when the collection looped over is a map.
2019-05-07 10:36:21 +03:00
4cf29bea36
MXS-2463 Rename m_nodes to m_nodes_by_id
...
Makes it obvious it is a map.
2019-05-07 10:36:21 +03:00
3c9ad0e452
MXS-2463 Ignore persisted data if bootstraps have changed
...
If there have been any changes in the bootstrap servers specified
for the Clustrix monitor, then the persistent connection information
is not used.
Otherwise, if the bootstrap server is changed and inaccessible, we
may connect to another cluster than the intended one.
2019-05-07 10:36:21 +03:00
ff33453e1a
MXS-2463 Prepare for another set of queries
...
Persisted information about dynamic nodes must be used only if
the bootrap information has not been changed, as otherwise we risk
using information that is not valid.
2019-05-07 10:36:21 +03:00
a9c1a8d18a
Don't build dependencies with newer compiler
2019-05-07 10:02:37 +03:00
8482337333
Merge branch '2.3' into develop
2019-05-06 20:16:02 +03:00
4e3a581c94
Merge branch '2.2' into 2.3
2019-05-06 20:09:37 +03:00
