Uploaded image for project: 'Red Hat Data Grid'
  1. Red Hat Data Grid
  2. JDG-3209

Massive Expiration can not be handled by the reaper without backlog and cause OutOfMemory

    XMLWordPrintable

Details

    • Bug
    • Status: Closed (View Workflow)
    • Critical
    • Resolution: Duplicate Issue
    • JDG 7.3 GA
    • None
    • None
    • None

    Description

      A use case where a huge number of entries are added and expire in a short period the reaper is not able to handle that and will have increasing backlog which cause OOM error.

      The TRACE messages within the logfiles are
      TRACE [org.infinispan.expiration.impl.ClusterExpirationManager] (pool-9-thread-1) Purging data container of expired entries
      TRACE [org.infinispan.expiration.impl.ClusterExpirationManager] (pool-9-thread-1) Submitting expiration removal for key WrappedByteArray

      {...}

      which had lifespan of 480000
      ..... 1,900,000 more
      TRACE [org.infinispan.expiration.impl.ClusterExpirationManager] (pool-9-thread-1) Purging data container completed in 15.45 minutes

      As shown the reaper run for >15 minutes but the lifespan is set to 480sec==8min so there is a massive overlap and the reaper is not able to delete the entries fast enough from the cache

      Attachments

        Issue Links

          Activity

            People

              wburns@redhat.com Will Burns
              rhn-support-wfink Wolf-Dieter Fink
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: