Distributed timers tests on OpenShift, with remote infinispan cluster are now failing regularly against WFLY 35 Beta.
Basically, once a timer executor is stopped ungracefully, no other instances take over in a reasonable time, and the timer stops ticking until a pod resembling the original executor is restarted and ready.
The tests start failing on one assertion that verifies that the recorded ticks number equals the one expected to occur within a given time window, currently reported as lower, instead - e.g.: expected 9, actual 4.
I guess this is due to the changes in https://issues.redhat.com/browse/WFLY-19419, as discussed above, and should be solved by https://issues.redhat.com/browse/WFLY-19550, and this JBEAP is logged since the current behavior in WildFly 35 Beta is different from the EAP 8 behavior, although it will be the base for the next EAP 8.1 release.