🚀 Release 1.19.5 (2026-11-12)

🗒 NOTES

StackGres 1.19.5 is out! 🐛 🔫 🎊 🍻

This release fixes the rollout of an SGCluster, performed by an SGDbOps of type restart, securityUpgrade or
minorVersionUpgrade or by the update strategy of the cluster, that could delete the primary Pod while the replica
Pod was being re-created, leaving the cluster without any available instance. It also fixes the data of the
instances of a cluster with a backup using volume snapshot being moved away at every start, leaving the cluster
without a leader when all its Pods started together. Finally, every change of the credentials of an SGCluster
is now applied without any manual step and without restarting its Pods, including the change of the username
of the superuser, of the replication user and of the monitor user.

So, what you are waiting for to try this release and have a look to the future of StackGres!

✨ NEW FEATURES AND CHANGES

Nothing new here! 👀

Web Console

Nothing new here! 👀

🐛 FIXES

• The rollout of an SGCluster deleted the primary Pod, without performing any switchover, while the replica Pod it had just deleted was still terminating or had not been re-created yet, leaving the cluster without any available instance (#3298)
  
• Once a backup using volume snapshot was available, the patroni container of an SGCluster moved its data directory away at every start unless the data had been restored from a volume snapshot, so a cluster whose Pods all started together (like after a rollout) waited forever for a leader to bootstrap (#3299)
  
• The description of the credentials of the users in the SGCluster and SGShardedCluster CRDs said that changing them required to modify the database users by hand, while the operator applies the password changes, and it did not describe which changes require a restart (#3302)
  
• Patroni kept using the previous credentials of the superuser, of the replication user and of its REST API after they were changed in an SGCluster, until its Pods were restarted, so a former primary could not rewind, the replicas could not reconnect and the requests to the Patroni REST API were rejected (#3301)
  
• Changing the username of the superuser or of the replication user of an SGCluster did not create the user, and blocked the other credentials changes and all the managed SQL scripts (#3304)
  
• A monitor user with a username other than monitor did not get the grants on the functions used by the Postgres exporter, so some metrics were not collected (#3305)
  
• Changing the username of the monitor user of an SGCluster required to restart its Pods to be applied to the Postgres exporter (#3306)
  
• The changed credentials of an SGCluster were applied to Patroni only at the next reconciliation of the cluster controller, up to a minute later, while the requests to the Patroni REST API already used the new password (#3307)
  
• The validating webhooks installed by the operator also validated the deletion of an SGCluster, while the ones of the operator Helm chart did not, so a tool that applies the chart, like Argo CD, reported the ValidatingWebhookConfiguration as out of sync after every restart of the operator. Since no validation acts on the deletion of a custom resource, the validating webhooks no longer validate deletions, which no longer fail while the operator is not available (#3308)
  

Web Console

Nothing new here! 👀

🚧 KNOWN ISSUES

• Backups may be restored with inconsistencies when performed with a Postgres instance running on a different architecture (#1539)
  

🆙 UPGRADE

To upgrade from a previous installation of the StackGres operator's helm chart you will have to upgrade the helm chart release.
For more detailed information please refer to our documentation.

To upgrade StackGres operator's (upgrade only works starting from 1.1 version or above) helm chart issue the following commands (replace namespace and release name if you used something different):

helm upgrade -n "stackgres" "stackgres-operator" https://stackgres.io/downloads/stackgres-k8s/stackgres/1.19.5/helm/stackgres-operator.tgz

IMPORTANT: This release is incompatible with previous alpha or beta versions. Upgrading from those versions will require uninstalling completely StackGres including all clusters and StackGres CRDs (those in stackgres.io group) first.

Thank you for all the issues created, ideas, and code contributions by the StackGres Community!

🔀 FULL LIST OF COMMITS