Michal Hajas

@pruivo I tested this today and I can still reproduce the failure. Also, the branch you linked is incorrect. Here is a detailed description of what is happening and how...

I can confirm the behavior described here: https://github.com/keycloak/keycloak-benchmark/issues/730 is still reproducible with this PR. My rosa-cross-dc config: ``` # GENERIC TIMEOUT=1200 # 20 minutes in seconds - 300 may be...

@pruivo It is described here: https://github.com/keycloak/keycloak-benchmark/issues/730#issue-2167195691 There is also a `How to Reproduce?` section.

This can be relaxed/made much less probable if we execute the updates with backoff on the Keycloak side. But we should probably first investigate how both sites may succeed with...

This needs to be revisited for A/A mode with persistent sessions. A test case should be added to the functional testsuite.

Do you suggest removing the option? Where we would then document how to set it for deployments with less memory? Also, if you suggest removing that option we should move...

There are two options 1. Use HotRod protocol to access the remote store with JBossMarshalling 2. Use Dataset provider functionality to access remote caches: https://github.com/keycloak/keycloak-benchmark/blob/main/dataset/src/main/java/org/keycloak/benchmark/cache/RemoteCacheResource.java Option 2. is better, although...

Sure. Yes, this is still necessary in my opinion. The issue here is that the CrossDC testsuite accesses Infinispan caches using ISPN's Rest API. This works well for some simple...

Revert https://github.com/keycloak/keycloak-benchmark/commit/b0a408a3f88104e4671ceabd1b3811fb026ef4f6 when this is resolved

We need to take Undertow and model tests into account if we decide to get rid of `persistence.xml`.