Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-4317

regular create and delete of buckets can lead to crash

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: 1.7.1
    • Fix Version/s: 2.0-beta
    • Component/s: couchbase-bucket
    • Security Level: Public
    • Labels:
      None

      Description

      Bug reporter states that creating 5 new buckets, with new names once an hour and deleting the 5 buckets created in the previous hour causes membase to "go down". See info in http://www.couchbase.org/forums/thread/membase-delete-bucket-doesnt-actually-free-ram-being-used

      No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

        Hide
        farshid Farshid Ghods (Inactive) added a comment - - edited

        we do create/delete buckets more than 40 times during our BAT test run and we dont see the memcached crash anymore.

        1- it can be the number of buckets the user is creating.
        2- in windows the max number of buckets the user could create in 2.0 dp2 was i think 4 or 5.
        3- memcached crash can also happen if user tries to create/delete buckets which have lot of data ( bucket-engine race condition which we have open bugs for )

        I will take a look at test_bucket_max test and see if its passing or failing.

        Show
        farshid Farshid Ghods (Inactive) added a comment - - edited we do create/delete buckets more than 40 times during our BAT test run and we dont see the memcached crash anymore. 1- it can be the number of buckets the user is creating. 2- in windows the max number of buckets the user could create in 2.0 dp2 was i think 4 or 5. 3- memcached crash can also happen if user tries to create/delete buckets which have lot of data ( bucket-engine race condition which we have open bugs for ) I will take a look at test_bucket_max test and see if its passing or failing.
        Hide
        farshid Farshid Ghods (Inactive) added a comment -

        Keith,

        can you please rerun this test 10 times and see if you see memcached crash

        createbuckettests.CreateMembaseBucketsTests.test_max_buckets

        Show
        farshid Farshid Ghods (Inactive) added a comment - Keith, can you please rerun this test 10 times and see if you see memcached crash createbuckettests.CreateMembaseBucketsTests.test_max_buckets
        Hide
        ingenthr Matt Ingenthron added a comment -

        Please read the comments in the forum posting. I think it's less about creating/deleting the same bucket, and more about possible resources not freed when creating new buckets even though the old ones have been deleted. The bug reporter says they're able to hang the system by destroying the last five and creating a new five buckets once an hour.

        It may be possible to do the same thing without the hour wait.

        Show
        ingenthr Matt Ingenthron added a comment - Please read the comments in the forum posting. I think it's less about creating/deleting the same bucket, and more about possible resources not freed when creating new buckets even though the old ones have been deleted. The bug reporter says they're able to hang the system by destroying the last five and creating a new five buckets once an hour. It may be possible to do the same thing without the hour wait.
        Hide
        keith Keith Batten (Inactive) added a comment -

        Our automated tests do delete/create quite a few buckets (apart from default, the bucket name is random though.) I'll look into loading a few buckets into DGM, then deleting and creating new ones, though I have never seen this cause problems before.

        Show
        keith Keith Batten (Inactive) added a comment - Our automated tests do delete/create quite a few buckets (apart from default, the bucket name is random though.) I'll look into loading a few buckets into DGM, then deleting and creating new ones, though I have never seen this cause problems before.
        Hide
        trond Trond Norbye added a comment -

        We've refactored the code path for bucket deletion to eliminate the potential race condition. Feel free to reopen the bug report if you still see the problem

        Show
        trond Trond Norbye added a comment - We've refactored the code path for bucket deletion to eliminate the potential race condition. Feel free to reopen the bug report if you still see the problem

          People

          • Assignee:
            Unassigned
            Reporter:
            ingenthr Matt Ingenthron
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Gerrit Reviews

              There are no open Gerrit changes