[Percona Community Forum] Summary

No dark-mode styles — this email renders the same in a dark client.

A brief summary of [Percona Community Forum][1] since 2026-10-26 02:27:08 UTC

7 New Topics
15 New Users


  
  ### Popular Topics

    [Percona Operator for MongoDB verison 1.23.1 upgrade issue][2]

      Description:
With an Openshift deployment of Percona Operator for MongoDB v 1.23.0 and the matching v1.23.0 psmdb-db helm chart. The operator was upgraded to v1.23.1. 
So the crVersion is 1.23.0 in the helm chart. 
The images in the psmdb-db chart were referencing the sha256 images in a private Red Hat Quay repository. 
The initContainer image was not overriding what is in the chart which is commented out by default. 
When the mongodb pod restarted an event/error occurred which stated that the image with tag 1.23.0 could not be parsed. 
Failed to apply default image tag “registry.connect.redhat.com/percona/percona-server-mongodb-operator@sha256:1.23.0”: couldn’t parse image name “registry.connect.redhat.com/percona/percona-server-mongodb-operator@sha256:1.23.0”: invalid reference format 
Previously no issue here with Operator v1.23.0. 
Steps to Reproduce:
Percona Operator for MongoDB v1.23.0 upgraded to v1.23.1 
Nothing yet done to Percona Server for Mongodb (psmdb-db) helm chart insta...

    
    [PXC Upgrade Management][3]

      Hello Percona Team, 
We are building a Database-as-a-Service (DBaaS) platform using Percona XtraDB Cluster (MySQL) and the Percona Kubernetes Operator. 
Our architecture consists of approximately 1,000 customer environments, where each customer has a dedicated three-node RKE2 Kubernetes cluster running on Apache CloudStack virtual machines. 
Current Architecture: 


Infrastructure: Apache CloudStack 


Kubernetes: RKE2 


Database: Percona XtraDB Cluster (MySQL) 


Database Management: Percona Kubernetes Operator 


Container Registry: Private Harbor registry 


Deployment: Dedicated Kubernetes cluster for each customer 


Scale: Approximately 1,000 independent customer clusters 


We are looking for a production-ready approach to centrally manage Percona XtraDB Cluster and Operator version upgrades across all customer environments without manually accessing individual Kubernetes clusters. 
We would appreciate your guidance on the following questions: 
1. Centralized Upgrade Management ...

    
    [Getting Started with Percona PostgreSQL Operator | Percona Community][4]

      I’m a long term user of MySQL with both asynchronous replication and synchronous with PXC. For a while now this has been on Kubernetes with either Helm deployments or more recently operator deployments. I felt it was time to dip my toe into the world of PostgreSQL as it’s become more and more popular (or certainly appears to be). 

This is a companion discussion topic for the original entry at https://percona.community/blog/2026/10/29/getting-started-with-percona-postgresql-operator/

    
    [PostgreSQL HA: Commit Durability and Consistent Reads During Failover][5]

      Hello PostgreSQL community, 
I’m researching general PostgreSQL high-availability and transaction-durability behavior and would appreciate guidance on the following: 


Acknowledged commits: Under which PostgreSQL replication and failover configurations can a successfully acknowledged transaction still be lost? 


Uncertain COMMIT outcomes: If a database connection drops before the client receives a COMMIT acknowledgment, what is the recommended way to determine the transaction’s outcome safely? 


Consistency after failover: What approaches prevent stale reads from replicas or newly promoted primary nodes when authoritative transaction state is required? 


Database unavailability: What general application safeguards are recommended when security-sensitive operations must be denied because authoritative database state cannot be verified? 


Are there particular PostgreSQL documentation references, Patroni configurations, or open-source testing tools you would recommend? 
Thank you for...

    
    [Catch Lost Updates Before Release: A Two-Client PostgreSQL … | Percona Community][6]

      Two successful reservations of three units should reduce stock from ten to four. A read-calculate-write operation can instead leave seven: both clients overwrite the row with a value calculated from an earlier read. That mismatch is what this lab checks. The release check is an invariant: 

This is a companion discussion topic for the original entry at https://percona.community/blog/2026/10/27/catch-lost-updates-before-release/

    

  




**New for you**

* [ Community Project: Open-Source MCP Server for PMM 2 & 3][7] - 1 - [PMM unofficial Dashboards and Plugins]
* [From Cake Pops to Valkey Pong: Finding My Place in the Open … | Percona Community][8] - 1 - [Community website]



[1]: https://forums.percona.com/
[2]: https://forums.percona.com/t/percona-operator-for-mongodb-verison-1-23-1-upgrade-issue/41387
[3]: https://forums.percona.com/t/pxc-upgrade-management/41388
[4]: https://forums.percona.com/t/getting-started-with-percona-postgresql-operator-percona-community/41381
[5]: https://forums.percona.com/t/postgresql-ha-commit-durability-and-consistent-reads-during-failover/41389
[6]: https://forums.percona.com/t/catch-lost-updates-before-release-a-two-client-postgresql-percona-community/41378
[7]: https://forums.percona.com/t/community-project-open-source-mcp-server-for-pmm-2-3/41391
[8]: https://forums.percona.com/t/from-cake-pops-to-valkey-pong-finding-my-place-in-the-open-percona-community/41384
[9]: https://forums.percona.com/



  This summary is sent from [Percona Community Forum][9] when we haven't seen you in a while. Change [your email settings][10], or [click here][11] to unsubscribe.

[10]: https://forums.percona.com/my/preferences/emails
[11]: https://forums.percona.com/email/unsubscribe/0214c62fd13b9eac51e6fa4d0443334498eb7449c1be9719c2bca6aceb8ebdd5

Accession

nse/percona/2026-11-02/5854

From

Percona Community Forum · notifications@percona1.discoursemail.com

Sent

2 November 2026 · 02:00 UTC