This is an 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.
Information about the wsrep sequence number has been added to
Percona XtraDB Cluster can now be bootstrapped with the new
bootstrap-pxcoption in the init script.
Percona XtraDB Cluster has implemented new Python version of the clustercheck script.
Percona XtraDB Cluster now has better integration with Percona XtraBackup 2.1 by implementing new
xbstreamoption for xtrabackup State Snapshot Transfer method.
Fixed the packaging issues that caused Percona XtraDB Cluster to conflict with MySQL libraries when performing fresh CentOS installation. Bug fixed #1183669.
Fixed the package conflict for
percona-xtradb-cluster-client-5.5package. Bug fixed #1172621.
RPMpackage dependences. Bug fixed #1050654.
mysql_tzinfo_to_sqlscript failed with Percona XtraDB Cluster since MyISAM DDL/DML were not consistently replicated. Fixed by enabling session variable
wsrep_replicate_myisamin the generator file. Bug fixed #1161432.
Startup script on Debian and Ubuntu was failing prematurely during
SST. Bug fixed #1099428.
Due to Debian way of stopping/reloading/status checking, there was failure after SST since the machine-specific Debian maintenance password didn’t work. Fixes by using signals instead of
mysqladminas it is done in CentOS. Bug fixed #1098360.
ib_logfile*files which the SST after that would remove. Bug fixed #1191767.
Percona XtraDB Cluster improved status visibility in the
processliston other nodes when provider is locked with
FLUSH TABLES WITH READ LOCKon one of them. Bug fixed #1109341.
Percona Server Crash-Resistant Replication was causing conflicts with
wsrep_crash_recoveryon a node which is a slave to an async master in standard MySQL replication. Bugs fixed #1182441 and #1180802.
FLUSH TABLES WITH READ LOCKdidn’t behave as expected, it had different behavior than the upstream MySQL version. Bug fixed #1170821.
When doing cascading Foreign Key processing Percona XtraDB Cluster was doing unnecessary Foreign Key checks, which are needed only for the purpose of getting Foreign Key parent reference for child table modifications. Bug fixed #1134892.
High priority threads (slave threads, replaying threads, total order isolation threads) were not protected from kill signal and attempt to kill high priority thread could cause server to hang. Bug fixed #1088267.
Galera provider may deadlock if applier thread is still executing
apply_trx()while processing commit, it could cause corresponding transaction to be purged from cert index.
Previous Percona XtraDB Cluster release introduced the regression in the RPM packaging that prevented the server from restarting following an upgrade. Bug fixed #1168032.
Other bug fixes: bug fixed #1187526, bug fixed #1133266, bug fixed #1079688, bug fixed #1177211, bug fixed #1166423, bug fixed #1084573, bug fixed #1069829, bug fixed #1012138, bug fixed #1170066, bug fixed #1170706, bug fixed #1182187, bug fixed #1183530, bug fixed #1183997, bug fixed #1184034, bug fixed #1191763, bug fixed #1069829.
Based on Percona Server 5.5.31-30.3 including all the bug fixes in it and on Codership wsrep API 5.5.31-23.7.5, Percona XtraDB Cluster 5.5.31-23.7.5 is now the current stable release. All of Percona‘s software is open-source and free.
For general inquiries, please send us your question and someone will contact you.