If you want to control when a RabbitMQ cluster gets updated, make sure to pause reconciliation before upgrading the cluster-operator.
After upgrading the cluster-operator, resume reconciliation whenever it's safe to update the RabbitMQ cluster.
See Pause reconciliation for a RabbitMQ cluster.
What's Changed
- Add --rabbitmq-version flag to kubectl-rabbitmq create by @Zerpet in #2156
- Add CLAUDE.md by @Zerpet in #2160
- Bump kind, k8s, and olm versions. by @MirahImage in #2161
- Fix system tests for RabbitMQ 3.13.x by @Zerpet in #2162
- Combine both gomod groups to avoid cross-dependent PRs. by @MirahImage in #2167
- Update startupProbe to use HTTP health endpoint, retain legacy option. by @MirahImage in #2171
- Pass CA to openssl to fix TLS example test. by @MirahImage in #2172
- feat: move defaulting logic to mutating webhook by @amaanx86 in #2151
- Rename webhook resources with cluster-operator- prefix by @Zerpet in #2177
- Ignore local Claude files by @Zerpet in #2178
- feat: integrate mutating webhook into OLM package by @Zerpet in #2174
- Fix CI certificate deployment validation. by @MirahImage in #2180
- feat(olm): add ValidatingAdmissionWebhook to CSV template by @Zerpet in #2184
- Bump Go deps by @Zerpet in #2185
New Contributors
Full Changelog: v2.21.1...v2.22.0