Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-4807

Set view group craches with reason {noproc, {gen_server,call, [<0.4370.0>,{drop,<0.4362.0>},infinity]}}

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Blocker
    • Resolution: Fixed
    • Affects Version/s: 2.0-developer-preview-4
    • Fix Version/s: 2.0-developer-preview-4
    • Component/s: None
    • Security Level: Public
    • Labels:
      None

      Description

      I was running a test against 4 node cluster (set up by cluster_run):

      python2 ./testrunner -i b/resources/dev-4-nodes.ini -t viewtests.ViewBasicTests.test_load_data_get_view_x_mins_y_design_docs,num-design-docs=25,load-time=10,run-view-time=2,num-tries=100

      After some time set view group crashed with this error:

      =========================CRASH REPORT=========================
      crasher:
      initial call: couch_set_view_group:init/1
      pid: <0.4362.0>
      registered_name: []
      exception exit: {noproc,
      {gen_server,call,
      [<0.4370.0>,

      {drop,<0.4362.0>}

      ,infinity]}}
      in function couch_set_view_group:init/1
      ancestors: [<0.3972.0>,<0.3961.0>]
      messages: [

      {'EXIT',<0.4371.0>,normal}

      ]
      links: [<0.3972.0>]
      dictionary: [{random_seed,{26506,11927,3989}}]
      trap_exit: true
      status: running
      heap_size: 10946
      stack_size: 24
      reductions: 116161
      neighbours:

      Seems to be similar to MB-4686. Log is attached.

      1. log.bz2
        221 kB
        Aliaksey Artamonau
      No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

        Hide
        damien damien added a comment -

        I believe this bug is as easy at catching and ignoring any error in couchdb close, as what is happening is the storage file is being deleted as it is rebalanced to another node. What I'm unsure about is if there are places where we might still be using the db (not just closing it) when this happens. If so, we might need to rethink how file deletions occurs.

        Show
        damien damien added a comment - I believe this bug is as easy at catching and ignoring any error in couchdb close, as what is happening is the storage file is being deleted as it is rebalanced to another node. What I'm unsure about is if there are places where we might still be using the db (not just closing it) when this happens. If so, we might need to rethink how file deletions occurs.
        Hide
        Aliaksey Artamonau Aliaksey Artamonau added a comment -

        I believe that the cluster was in steady state when this happened (i.e. no recent rebalance). I didn't save all the log files. But I will double check by the log files attached to MB-4809.

        Show
        Aliaksey Artamonau Aliaksey Artamonau added a comment - I believe that the cluster was in steady state when this happened (i.e. no recent rebalance). I didn't save all the log files. But I will double check by the log files attached to MB-4809 .
        Hide
        Aliaksey Artamonau Aliaksey Artamonau added a comment -

        It seems that MB-4809 is a bit different. There it really happened right after rebalance. Here last rebalance happened 30 minutes before these errors. Stupid of me not to save all the diags.

        Show
        Aliaksey Artamonau Aliaksey Artamonau added a comment - It seems that MB-4809 is a bit different. There it really happened right after rebalance. Here last rebalance happened 30 minutes before these errors. Stupid of me not to save all the diags.
        Hide
        FilipeManana Filipe Manana (Inactive) added a comment -

        Fixed with several changes merged this week.

        Show
        FilipeManana Filipe Manana (Inactive) added a comment - Fixed with several changes merged this week.

          People

          • Assignee:
            FilipeManana Filipe Manana (Inactive)
            Reporter:
            Aliaksey Artamonau Aliaksey Artamonau
          • Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Gerrit Reviews

              There are no open Gerrit changes