Details

    • Type: Bug
    • Status: Resolved
    • Priority: High
    • Resolution: Cannot Reproduce
    • Affects Version/s: 5.2.0, 5.5.1
    • Fix Version/s: 6.6.0
    • Component/s: Database
    • Labels:
      None

      Description

      Problem:
      Deploying many large RPM artifacts/builds (40+) concurrently, or near concurrently will result in a number of 500 errors where Artifactory is unable to update metadata on all nodes:

      2017-10-29 21:17:08,892 [http-nio-8081-exec-170] [ERROR] (o.a.w.s.RepoFilter :224) - Sending HTTP error code 500: Lock on LockEntryId rpm-local:art-artifactory-powerpack-rpm-3.9.0.rpm-test.rpm not acquired in 120 seconds. Lock info: org.artifactory.storage.db.locks.provider.DbMapLockWrapper@7af7055.
      ...
      2017-10-29 21:21:30,654 [art-exec-442] [ERROR] (o.a.a.y.YumServiceImpl:198) - Could not save cache rpm metadata [rpm-local:art-jfrog-artifactory-pro-5.5.1.rpm-test.rpm]
      2017-10-29 21:23:30,657 [art-exec-442] [WARN ] (o.a.s.d.l.p.DbMapLockWrapper:46) - Could not acquire lock within 120 seconds. Couldn't acquire lock for: 120000 milliseconds
      2017-10-29 21:23:30,671 [art-exec-442] [ERROR] (o.j.m.y.w.YumRepoMetadataWriter:137) - Failed to serialize RPM metadata of 'art-jfrog-artifactory-pro-5.5.1.rpm-test.rpm'
      org.artifactory.concurrent.LockingException: Lock on LockEntryId rpm-local:art-jfrog-artifactory-pro-5.5.1.rpm-test.rpm not acquired in 120 seconds. Lock info: org.artifactory.storage.db.locks.provider.DbMapLockWrapper@13f066a4.

      20171028142737|880793|REQUEST|104.196.245.50|admin|PUT|/rpm-local/art-test-test-test-test-artifactory-powerpack-rpm-3.9.5.rpm-.rpm-.rpm-.rpm-.rpm-test.rpm|HTTP/1.1|500|118132446
      2017-10-28 14:27:37,818 [http-nio-8081-exec-136] [ERROR] (o.a.w.s.RepoFilter :224) - Sending HTTP error code 500: Failed to update the unique indices table

      2017-10-27 23:22:58 |599588|REQUEST|104.196.245.50|admin|PUT|/rpm-local/art-test-test-test-test-upload.sh-.rpm-.rpm-.rpm-.rpm-test.rpm|HTTP/1.1|500|118132446
      2017-10-27 23:22:58,101 [http-nio-8081-exec-175] [ERROR] (o.a.w.s.RepoFilter :224) - Sending HTTP error code 500: Failed to delete properties for node: 9280

      {
      "errors" : [

      Unknown macro: { "status" }

      ]
      }{
      "errors" : [

      Unknown macro: { "status" }

      Reproducing steps:
      1. Artifactory HA cluster on 5.5.1, MYSQL_DB
      2.Attempt deploying RPMs to Repository A:bitbucket_repo/bitbucket_branch/build_num/rpm_name/
      3. Copy RPMs from Repository A:bitbucket_repo/bitbucket_branch/build_num/rpm_name/ to Repository B:yum_repo_name/UUID/rpm_name/

      Used this script to upload:
      #!/bin/sh

      for run in 1 * 40
      do

      curl -u admin:password -T jfrog-artifactory-pro-5.4.2.rpm "http://mill.jfrog.info:34839/artifactory/rpm-local/art-$run-test.rpm" &

      done

      Support added this plugin to expedite step 3: https://github.com/JFrogDev/artifactory-user-plugins/tree/6fd41c4032b5b731cc91fd833ed643434bc58bac/storage/createCopy
      With these properties:
      repository = ["rpm-local"] //if something is created here...
      repocopy = ["rpm-backup"] //copy it to the corresponding repo here

        Attachments

          Issue Links

            Activity

              People

              • Assignee:
                inbart Inbar Tal
                Reporter:
                loreny Loren Yeung
              • Votes:
                3 Vote for this issue
                Watchers:
                6 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: