Running ELK 5.6.5 with persistent queue enabled in logstash.yml running on centos 7.1
Run into following exception frequently
org.logstash.ackedqueue.io.AbstractByteBufferPageIO$PageIOInvalidVersionException: Expected page version=1 but found version=0
entire server running ELK is powered off instantaneously (or sometime given 5 seconds after SIGTERM) for High Availability testing. Only way to recover (restart logstash) is delete the queue and folder.
Is this issue been fixed or any workaround?
--pipeline.unsafe_shutdown with or without using this startup option didn't make any difference