Details

      Description

      I started getting periodic stack overflow errors which sometimes bring the service down completely. The backtrace is like this:

      Oct 28 18:21:48 wdqs1001 bash[2102]: Caused by: java.lang.StackOverflowError
      Oct 28 18:21:48 wdqs1001 bash[2102]: at java.util.concurrent.locks.ReentrantLock.lockInterruptibly(ReentrantLock.java:340)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.cache.HardReferenceQueueWithBatchingUpdates.add(HardReferenceQueueWithBatchingUpdates.java:490)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.AbstractHTree.doTouch(AbstractHTree.java:1631)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.AbstractHTree.touch(AbstractHTree.java:1486)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.DirectoryPage.checkLazyChild(DirectoryPage.java:471)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.DirectoryPage.getChild(DirectoryPage.java:370)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.DirectoryPage.getChildIfPresent(DirectoryPage.java:1791)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.htree.HTree.lookupAll(HTree.java:1323)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at com.bigdata.bop.join.HTreeHashJoinUtility$4.expand(HTreeHashJoinUtility.java:1959)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at cutthecrap.utils.striterators.Expanderator.getNext(Expanderator.java:59)
      Oct 28 18:21:48 wdqs1001 bash[2102]: at cutthecrap.utils.striterators.Expanderator.getNext(Expanderator.java:59)
      

      and then the last line is repeated a very large number of times. I'll try to see if it's linked to a particular query.

      1. BLZG-1890.patch
        2 kB
        Brad Bebee
      2. q1.rdf.gz
        400 kB
        stasmalyshev
      3. q2.rdf.gz
        0.8 kB
        stasmalyshev

        Activity

        Hide
        beebs Brad Bebee added a comment -

        BLZG-1890 patch for BLZG-3644.

        Show
        beebs Brad Bebee added a comment - BLZG-1890 patch for BLZG-3644 .
        Hide
        beebs Brad Bebee added a comment -

        It looks like this one BLZG-1890 is slated for 2.2.0 (though not exactly sure why it didn't roll into 2.1.x). The patch is BLZG-1890.patch.

        Internal reference is https://github.com/blazegraph/bigdata/commit/b3cd421232f226c7721e7d5faee39a350384c30b.

        stasmalyshev Can you confirm this fixes the issue?

        Show
        beebs Brad Bebee added a comment - It looks like this one BLZG-1890 is slated for 2.2.0 (though not exactly sure why it didn't roll into 2.1.x). The patch is BLZG-1890.patch . Internal reference is https://github.com/blazegraph/bigdata/commit/b3cd421232f226c7721e7d5faee39a350384c30b . stasmalyshev Can you confirm this fixes the issue?
        Hide
        beebs Brad Bebee added a comment -
        Show
        beebs Brad Bebee added a comment - stasmalyshev The commit is also available in the 2.2.0_RC branch. https://github.com/blazegraph/database/commit/b3cd421232f226c7721e7d5faee39a350384c30b
        Hide
        stasmalyshev stasmalyshev added a comment -

        Looks like this patch leads exactly to the code I've already tested, so it would work the same too.

        Show
        stasmalyshev stasmalyshev added a comment - Looks like this patch leads exactly to the code I've already tested, so it would work the same too.
        Hide
        beebs Brad Bebee added a comment -

        Cherry-picked into 2.1.5 RC BRANCH.

        Show
        beebs Brad Bebee added a comment - Cherry-picked into 2.1.5 RC BRANCH.

          People

          • Assignee:
            beebs Brad Bebee
            Reporter:
            stasmalyshev stasmalyshev
          • Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved: