Announcing The Release Of repmgr 4.0.5
Oxford, United Kingdom – May 3, 2018
2ndQuadrant today announced the release of repmgr 4.0.5. repmgr is the most popular tool for PostgreSQL replication and failover management. It enhances and complements the built-in replication capabilities in PostgreSQL, making it easy to set up and manage standby servers, check replication status, and perform administrative tasks such as switchover operations. The repmgrd daemon provides automatic failover capability to maximize database uptime, and can easily be integrated with other tools such as PgBouncer for high application availability.
This release is the first to provide packages via 2ndQuadrant’s new public RPM and APT repositories; for more details view the installation packages document here.
repmgr 4.0.5 includes usability enhancements related to pg_rewind usage, recovery.conf generation and, in repmgrd, handling of various corner-case situations. These enhancements are:
- Poll demoted primary after restart as a standby during a switchover operation
- Add configuration parameter config_directory
- Add sanity check if --upstream-node-id not supplied when executing repmgr standby register
- Enable pg_rewind to be used with PostgreSQL 9.3/9.4
- When generating replication connection strings, set dbname=replication if appropriate.
- Enable provision of archive_cleanup_command in recovery.conf
- Check actively for node to rejoin cluster
- Set connect_timeout=2 in repmgrd if not explicitly set, when pinging a server
- Various documentation improvements, with particular emphasis on the importance of setting appropriate service commands instead of relying on pg_ctl
This release contains the following bug fixes:
- Fix display of conninfo parsing error messages.
- Fix minimum accepted value for degraded_monitoring_timeout
- Fix superuser password handling
- Fix parsing of archive_ready_critical configuration file parameter
- Fix repmgr_cluster_crosscheck output
- Fix memory leaks in witness code
- Handle pg_ctl promote timeout in repmgrd
- Handle failover situation in repmgrd with only two nodes in the primary location, and at least one node in another location
- Prevent standby connection handle from going stale in repmgrd
For a detailed list of all changes, please read the full release notes here.
We highly recommend upgrading to repmgr 4.0.5 at the earliest opportunity available. For this release, a simple package upgrade from repmgr 4.0 – 4.0.4 is required; repmgrd (if running) should be restarted.
repmgr is developed and maintained by 2ndQuadrant – your diligent team of PostgreSQL experts.
repmgr is distributed under GPL v3.
For more information, please send an email to [email protected]