Skip to content

Releases: signal18/replication-manager

Replication Manager 2.0.0 Release Candidate 2

29 Jan 16:24
Compare
Choose a tag to compare

This is the 2nd release candidate of Replication Manager for MySQL/MariaDB/Percona series.

Changelog

  • Fixed several issues with Arbitrator server
  • RM will not overrule ProxySQL on read-only flag detections
  • Last RW server is now correctly set to master instead of standalone
  • ProxySQL can now be bootstrapped from the RM config
  • Maintenance toggle was broken and has been fixed
  • Fixed issues with slave down calling for replication checks
  • Clean up some unnecessarily verbose log messages
  • Fixed configuration typos
  • Preferred and Ignored servers are now shown in the dashboard
  • Graphs tab will be disabled if Graphite hasn't been configured
  • Fixed switchover to preferred master on CLI
  • Fixed Could not set replication channel when a db server has no binlog (#198)
  • Rewrote Dockerfile to work with 2.x series

Replication Manager 2.0.0-rc1

05 Jan 15:45
Compare
Choose a tag to compare

This is a release candidate version of Replication Manager 2.0.
Please refer to https://docs.signal18.io/ for configuration changes.

  • Replication Manager API
  • Rewritten MaxScale Proxy monitoring
  • Support of MySQL replication
  • Support of OpenSVC enterprise
  • New unified configuration syntax
  • Support for ProxySQL
  • Arbitrator in standalone package
  • Refactored internal logging
  • Improved HTTP Interface
  • Improved packaging
  • Multi Master Ring support
  • Consul Support
  • Maintenance Mode support
  • Failover using Pseudo GTID support
  • Rewritten HAProxy support
  • Shard Proxy (Spider) support
  • Sphinx support
  • Grafana integration
  • Refactored monitoring logic
  • Too many bug fixes to be listed here

Replication Manager v1.1.3 Maintenance Release

10 Nov 09:20
Compare
Choose a tag to compare

MariaDB/MySQL Replication Manager 1.1.2 release

04 Sep 15:44
Compare
Choose a tag to compare

This is a bug fix release to address issues with cascading failover and autorejoin not functioning.

MariaDB/MySQL Replication Manager 1.1.1 release

18 Aug 09:32
Compare
Choose a tag to compare

General Availability Release for replication-manager 1.1.1

Fixes:

  • Fix slave stuck in suspect mode
  • Fix gtid_slave_pos in dbhelper
  • Change wait apply relay log for GTID Position in MariaDB
  • Fix inserting fake transaction during failover
  • Fix arbitration issue where a SQLite transaction was left open
  • Fix MaxScale provisioning
  • Various logging fixes

Replication Manager 1.1.0 Preview 3

04 Jul 10:28
Compare
Choose a tag to compare
Pre-release
  • Fix issue when reintroducing slave
  • Fix error messages

replication-manager 1.1.0 Preview 2

26 Jun 14:36
Compare
Choose a tag to compare
Pre-release

Release notes

  • Added MySQL 5.7 GTID support and 5.6 positional support
  • Improved multi-MRM cluster
  • More support for Multi-Tiered Slaves
  • Multiple topologies can now be bootstrapped
  • Improved errors
  • Arbitrator now uses sqlite instead of mysql
  • Improved service files
  • Added MariaDB 10.2 Support
  • Included mysqlbinlog
  • Support for MariaDBShardProxy (Spider)
  • Too many bug fixes to be listed here
  • Improved documentation

replication-manager 1.1.0 Preview 1

20 Mar 11:47
Compare
Choose a tag to compare
Pre-release
  • Support for Multi Cluster
  • Rejoin Cluster via mysqldump
  • Rejoin Cluster via Flashback (MariaDB 10.2)
  • Improved testing framework
  • Auto-enforce mode
  • Maxscale mode improvements
  • Many small bugs corrected

Know issues

Http authentication should not be used causing memory leak if browser stay open
Standby server does not follow topology. Already fixed in dev

replication-manager 1.0.2 bugfix release

02 Feb 12:23
Compare
Choose a tag to compare

Release notes

  • Fix EL6 initscript issues

replication-manager 1.0.1 release

31 Jan 15:56
Compare
Choose a tag to compare

Release notes

  • Fixed a panic issue with configuration file not found or unreadable
  • Fixed version numbering issues
  • Added a warning vs. using http-auth

Known issues

  • http-auth currently leaks memory if a browser is open on the dashboard while it is activated