Repository navigation
Best Practices Guidance: Mitigate Maintenance Downtime #41690
Unanswered
cloud9architect
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Overview
It is known behavior using the latest release of Keycloak hosted under the Kubernetes platform that any updates made to the Keycloak image name or tag (including patches) will require the Operator to scale down the Statefulset as part of the redeployment operation.
See: https://www.keycloak.org/nightly/operator/rolling-updates
This process unfortunately introduces downtime, which can be frequent and problematic for organizations having high SLO requirements and/or little tolerance for service disruptions such as within the Healthcare IT Services vertical.
As outlined via the following epic, incremental steps are being made to support rolling updates for the above scenarios, and also eliminate breaking database changes during patch upgrades: #38881
This presents several questions in the interim regarding best practices and ways to mitigate downtime during maintenance.
Questions
Until #38881 is fully completed and released to GA, I would have the following questions below:
How are organizations that maintain high SLA/SLO requirements reducing downtime in Kubernetes that is triggered by the Keycloak image name or tag changing?
Are there any Kubernetes high availability architectures or release management strategies outside of obvious intra-cluster HA, active/active, and active/passive cluster designs (preferably hosted across the Microsoft Azure platform via AKS) that organizations have found successful in significantly reducing downtime during Keycloak updates?
Are there any recommended design strategies that greenfield Keycloak projects hosted in Azure Kubernetes Service (AKS) can employ that would better prepare for the future planned functionality of supporting rolling updates across patch versions?
Is there any additional functionality for Keycloak that is either GA or planned for release in the future outside of #38881 that can assist with increased reliability during maintenance and reduced downtime?
Thanks
All reactions