Percona is glad to announce the release of Percona XtraDB Cluster 5.5.34-23.7.6 on November 4, 2013. Binaries are available from the downloads area or from our software repositories.
Bugs Fixed:
- Fixed the issues with upgrade on Debian caused by
debian-start
script. Bug fixed #1112363. - Adding the
AUTO_INCREMENT
column to a table would result in data inconsistency because different nodes use differentauto_increment_offset
. Bug fixed #587170. LOAD DATA INFILE
was not replicating properly in cluster. Bug fixed #1206129.- After a
COMMIT
processing was interrupted (e.g. certification failure) it didn’t always release transactionalMDL
locks, which could lead to server deadlock while performingDDL
. Bug fixed #1212955. pc.checksum
option was enabled by default, although its purpose was just to help with debugging. Bug fixed #1229107.- Server startup isn’t interrupted if the
/var/lock/subsys/mysql
file was not cleaned up by the init script. Bug fixed #1231067. - State Snapshot Transfer cleanup of existing data directory is now configurable with a
cpat
parameter. Bug fixed #1231121. - Handler command and concurrent
DDL
could cause node hanging. The problem would happen when a handler session was aborted due to replication conflict. Bug fixed #1233353. - Command
service mysql status
was not producing valid exit codes, which could break the configuration management systems. Bug fixed #1237021. - Option
wsrep-new-cluster
could lead to server crash. Bug fixed #1238331. - No key information was provided for
DROP VIEW TOI
statement. Bug fixed #1240040. innodb_locks_unsafe_for_binlog
option is no longer checked because slave threads areREAD-COMMITTED
by default. Bug fixed #1240112.- if IO cache contained the
Query_log_event
, database pointer would be reset during transaction replay. This could lead to“1046 No database selected.”
error when next statement is issued. Bug fixed #1241760. - After test shutting down all the nodes indicate that all write set handlers were not released. Bug fixed #1244661.
MDL
locks were not released properly on slaves at commit time. Bug fixed #1245333.- Implemented number of thread pool scheduler fixes. Bug fixed #1240500.
- Running
FLUSH STATUS
would zero-up wsrep_cluster_size and wsrep_local_index variables. Bug fixed #1232789. - Fixed the dependency conflict between the
Percona-XtraDB-Cluster-shared
andPercona-Server-shared-compat
packages. Bug fixed #1084309. - Fixed the memory leak in the
wsrep_write_cache()
function. Bug fixed #1230160. - Xtrabackup SST Configuration implementation added in Percona XtraDB Cluster 5.5.33-23.7.6 has been renamed xtrabackup-v2, so wsrep_sst_method=xtrabackup will use xtrabackup implementation before 5.5.33-23.7.6 and will be compatible with older Percona XtraDB Cluster versions. Bug fixed #1228618.
- Support for SSL encryption for just the key and crt files as implemented in Galera can be enabled with
encrypt=3
option. This has been implemented in 5.5.34-23.7.6 for compatibility with Galera. Bug Fixed #1235244.
Other bugs fixed: bug fixed #1244741, bug fixed #1244667, bug fixed #1243150, bug fixed #1232890, bug fixed #999492, bug fixed #1245769, and bug fixed #1244100.
Based on Percona Server 5.5.34-32.0 including all the bug fixes in it, Galera Replicator and on Codership wsrep API 5.5.34-25.9, Percona XtraDB Cluster 5.5.34-23.7.6 is now the current stable release. Percona XtraDB Cluster 5.5.33-23.7.6 to 5.5.34-23.7.6 upgrade is strongly recommended if SST backward compatibility is a concern. All of Percona’s software is open-source and free.
This is a General Availability release. We did our best to eliminate bugs and problems during the testing release, but this is a software, so bugs are expected. If you encounter them, please report them to our bug tracking system.
Errata and the release notes for Percona XtraDB Cluster 5.5.34-23.7.6 are available in our online documentation.
Comment (1)
Currently we using percona xtraDB cluster, please let us know simple command for upgrade the cluster version without any downtime.
Comments are closed.
Use Percona's Technical Forum to ask any follow-up questions on this blog topic.