- Update the documentation to clarify that if patroni cannot be reached, we consider it's a configuration error that returns UNKNOWN (#77, reported by @MLyssens)
- cluster_has_replica now properly accounts for standby leaders (#72, reported by @MLyssens)
- Update the tests and the documentation to reflect that master is replaced by primary everywhere it's visible in Patroni. We didn't use the term so there is nothing to change in our code.
While fixing cluster_has_replica
, the definition of a healthy replica was
changed. It's more restrictive now, hence the jump from v1 to v2.
- In
cluster_node_count
, a healthy standby, sync replica or standby leaders cannot be "in archive recovery" because this service doesn't check for lag and timelines.
- Add the timeline in the
cluster_has_replica
perfstats. (#50) - Add a mention about shell completion support and shell versions in the doc. (#53)
- Add the leader type and whether it's archiving to the
cluster_has_leader
perfstats. (#58)
- Add compatibility with requests version 2.25 and higher.
- Fix what
cluster_has_replica
deems a healthy replica. (#50, reported by @mbanck) - Fix
cluster_has_replica
to display perfstats for replicas whenever it's possible (healthy or not). (#50) - Fix
cluster_has_leader
to correctly check for standby leaders. (#58, reported by @mbanck) - Fix
cluster_node_count
to correctly manage replication states. (#50, reported by @mbanck)
- Improve the documentation for
node_is_replica
. - Improve test coverage by running an HTTP server to fake the Patroni API (#55 by @dlax).
- Work around old pytest versions in type annotations in the test suite.
- Declare compatibility with click version 7.1 (or higher).
- In tests, work around nagiosplugin 1.3.2 not properly handling stdout redirection.
Check patroni is now tagged as Production/Stable.
- Add
sync_standby
as a valid replica type forcluster_has_replica
. (contributed by @mattpoel) - Add info and options (
--sync-warning
and--sync-critical
) about sync replica tocluster_has_replica
. - Add a new service
cluster_has_scheduled_action
to warn of any scheduled switchover or restart. - Add options to
node_is_replica
to check specifically for a synchronous (--is-sync
) or asynchronous node (--is-async
). - Add
standby-leader
as a valid leader type forcluster_has_leader
. - Add a new service
node_is_leader
to check if a node is a leader (which includes standby leader nodes)
- Fix the
node_is_alive
check. (#31) - Fix the
cluster_has_replica
andcluster_node_count
checks to account for the new replica statestreaming
introduced in v3.0.4 (#28, reported by @log1-c)
- Create CHANGELOG.md
- Add tests for the output of the scripts in addition to the return code
- Documentation in CONTRIBUTING.md
- Add a
--save
option when state files are used - Modify
-e/--endpoints
to allow a comma separated list of endpoints (#21, reported by @lihnjo) - Use requests instead of urllib3 (with extensive help from @dlax)
- Change the way logging is handled (with extensive help from @dlax)
- Reverse the test for
node_is_pending
- SSL handling
- Several doc Fix and Updates
- Use spellcheck and isort
- Remove tests for python 3.6
- Add python tests for python 3.11
The initial release covers the following checks :
- check a cluster for
- configuration change
- presence of a leader
- presence of a replica
- maintenance status
- check a node for
- liveness
- pending restart status
- primary status
- replica status
- tl change
- patroni version