- Percona XtraDB Cluster has implemented initial implementation of weighted quorum. Weight for node can be assigned via pc.weight option in the wsrep_provider_options variable. Accepted values are in the range [0, 255] (inclusive). Quorum is computed using weighted sum over group members.
- Percona XtraDB Clusterbinary will now be bundled with the
RPM/debpackages, this library will be available for download from our repositories. Benchmark showing the impact of memory allocators on MySQL performance can be found in this blogpost.
- This release of Percona XtraDB Cluster has fixed number of foreign key and packaging bugs.
- Fixed yum dependencies that were causing conflicts in CentOS 6.3 during installation. Bug fixed #1031427 (Ignacio Nin).
- In case the Percona XtraDB Cluster was built from the source rpm, wsrep revision information would be missing. Bug fixed #1128906 (Alexey Bychko).
- The method of generating md5 digest over tuples in a table with no primary key was not deterministic which could lead to a node failure. Bug fixed #1019473 (Seppo Jaakola).
- Percona XtraDB Cluster was built with YaSSL which could cause some of the programs that use it to crash. Fixed by building packages with OpenSSL support rather than the bundled YaSSL library. Bug fixed #1104977 (Raghavendra D Prabhu).
- Clustercheck script would hang in case the MySQL server on a node is hung. As a consequence clustercheck script would never fail-over that server. Bug fixed #1035927 (Raghavendra D Prabhu).
- High values in variables evs.send_window and evs.user_send_window could trigger cluster crash under high load. Bug fixed #1080539 (Teemu Ollakka).
- Standard MySQL port would be used when port number wasn’t explicitly defined in the wsrep_node_incoming_address. Bug fixed #1082406 (Alex Yurchenko).
- Dropping a non-existing temporary table would be replicated when TOI was used in wsrep_OSU_method variable. This bug was fixed for the case when
DROP TEMPORARY TABLEstatement was used, but it will still replicate in case
DROP TABLEstatement is used on a temporary table. Bug fixed #1084702 (Seppo Jaakola).
- In case two nodes in a 3-node cluster had to abort due to inconsistency, one wouldn’t correctly notify the surviving node which would lead to surviving node to loose the primary component and cause subsequent downtime. Bug fixed #1108165 (Alex Yurchenko).
- In some cases non-uniform foreign key reference could cause a slave crash. Fixed by using primary key of the child table when appending exclusive key for cascading delete operation. Bug fixed #1089490 (Seppo Jaakola).
- Parallel applying would fail in case mixed
VARCHARcolumns would be used in foreign key definitions. Bug fixed #1100496 (Seppo Jaakola).
- Debian packages included the old version of innotop. Fixed by removing innotop and its
InnoDBParserPerl package from source and Debian installation. Bug fixed #1032139 (Alexey Bychko).
mysqld_safescript would fail to retrieve the Galera replication position on Ubuntu 10.04, because the different shell was used. Bug fixed #1108431 (Alex Yurchenko).
- Cascading foreign key constraint could lead to unresolved replication conflict and leave a slave hanging. Bug fixed #1130888 (Seppo Jaakola).
- If MySQL replication threads were started before running wsrep recovery, this would lead to memory corruption and server crash. Bug fixed #1132974 (Seppo Jaakola).
- Conflicting prepared statements in multi-master use case could cause node to hang. This was happening due to prepared statement execution loop, which does not honor wsrep status codes correctly. Bug fixed #1144911 (Seppo Jaakola).
- State Snapshot Transfer with Xtrabackup would fail if the
tmpdirwas specified more than once in the MySQL configuration file (
my.cnf). Bugs fixed #1160047 and #1086978 (Raghavendra D Prabhu).
- Donor node would run XtraBackup indefinitely when xtrabackup
tmpdirwas set up on
tmpfs. Bug fixed #1086978 (Alex Yurchenko).
- Issues with compiling Galera on the
ARMarchitecture has been fixed. Bug fixed #1133047 (Alex Yurchenko).
- Upstream bugfix for bug #59354 triggered a regression that could cause transaction conflicts. Bug fixed #1158221 (Seppo Jaakola).
- Galera builds would fail when they were built with the new
boostlibrary. Bug fixed #1131736 (Alex Yurchenko).
lost+foundwasn’t included in the rsync SST filter, this caused the SST failure due to insufficient privileges. Fixed by excluding
lost+foundfolder if found. Bug fixed #1154095 (Alex Yurchenko).
- If variable innodb_thread_concurrency has been defined to throttle InnoDB access, and work load contained DDL statements, a cluster node could remain hanging for unresolved MDL conflict. Fixed by adding a new method to cancel a thread waiting for InnoDB concurrency. Bug fixed #1155183 (Seppo Jaakola).
- Handling of the network issues in Galera has been improved. Bug fixed #1153727 (Teemu Ollakka).
- Fixed the wrong path in the
/etc/xinetd.d/mysqlchkscript. Bugs fixed #1000761 and #1132934 (Raghavendra D Prabhu).
- When upgrading the
/usr/bin/clustercheckscript would get overwritten, and any changes (such as username and password) would be lost. Bug fixed #1158443 (Raghavendra D Prabhu).
- In case
CREATE TABLE AS SELECTstatement was running in parallel with the DDL statement on the selected table, in some cases first statement could be left hanging. Bug fixed #1164893 (Seppo Jaakola).
- Galera builds would fail when
gcc4.8 was used. Bug fixed #1164992 (Alex Yurchenko).
Percona-XtraDB-Cluster-galerapackage version number didn’t match the wsrep_provider_version one. Bug fixed #1111672 (Alexey Bychko).
rpmdebug build was available for Percona XtraDB Cluster, fixed by providing the
debdebug build as well. Bug fixed #1096123 (Ignacio Nin).
Based on Percona Server 5.5.30-30.2 including all the bug fixes in it and on Codership wsrep API 5.5.30-23.7.4, Percona XtraDB Cluster 5.5.30-23.7.4 is now the current stable release. All of Percona’s software is open-source and free. Release notes for Percona XtraDB Cluster 5.5.30-23.7.4 are available in our online documentation.
We did our best to eliminate bugs and problems, but this is a software, so bugs are expected. If you encounter them, please report them to our bug tracking system.
UPDATE[18-04-2013]: There was a RPM packaging regression introduced with the fix for bug #710799. This regression only affected clean RPM installations and not upgrades. We have pushed the fixed packages to the repositories. Bug fixed #1170024.