I've got a stand-alone application leveraging the following stack: Hibernate 5.6 Infinispan 13.0.15 JGroups 4.2.22
I'm using Infinispan as a second level cache for Hibernate -- I've noticed that when I start up my second application node I get the following error:
ISPN000136: Error executing command ReadWriteKeyCommand on Cache 'clientDeviceNotificationTimeRangeCache', writing keys [com.xxx.ClientDeviceNotificationTimeRangeVO#ClientDeviceNotificationTimeRangePrimaryKey [clientId=5041, dayOfWeek=1]]| java.util.NoSuchElementException: No value present
This causes the primary node to throw an accompanying error:
org.infinispan.util.concurrent.TimeoutException: ISPN000299: Unable to acquire lock after 0 milliseconds for key com.xxx.ClientDeviceNotificationTimeRangeVO#ClientDeviceNotificationTimeRangePrimaryKey [clientId=5041, dayOfWeek=5] and requestor CommandInvocation:localhost-54954:4420. Lock is held by null
It seems like the cache is replicating objects to the secondary node before the secondary node is actually fully up and available -- strangely this worked perfectly fine in the 8.0 series .. any obvious thing I might be overlooking here? some new setting or flag I am omitting? the only other oddity I can see here is what is setting a 0 milliseconds lock timeout for a key? that doesn't seem right ..