+++ This bug was initially created as a clone of Bug #1208520 +++ Description of problem: When the glusterfind pre command is run with change-detector as 'brickfind' it works fine. When it is run with 'changelog' it errors out saying change detection failed. Version-Release number of selected component (if applicable): Gluster 3.7 upstream nightly glusterfs-3.7dev-0.852.git3feaf16.el6.x86_64 How reproducible: Always Steps to Reproduce: 1. Have a gluster volume, say 'ozone' and create a glusterfind session glusterfind create <sessionName> <volumeName> 2. Run the glusterfind pre command with change-detector as 'changelog' glusterfind pre <sessionName> <volumeName> /tmp/outfile.txt --change-detector=changelog Actual results: Step2 should succeed. The specified output file 'outfile.txt' should get updated with the list of changed files, from the last change Expected results: Step 2 errors out saying: Change detection failed Additional info: Could not get much info from the logs as well. It just says: Change detection failed. [2015-04-02 23:37:53,689] ERROR [utils - 115:fail] - 10.70.43.48 - Change detection failed [2015-04-02 23:37:53,733] ERROR [utils - 115:fail] - 10.70.43.48 - Change detection failed [2015-04-02 23:37:53,745] ERROR [utils - 115:fail] - 10.70.43.48 - Change detection failed [2015-04-02 23:37:54,867] ERROR [utils - 115:fail] - 10.70.42.147 - Change detection failed [2015-04-02 23:37:54,867] ERROR [utils - 115:fail] - 10.70.42.147 - Change detection failed [2015-04-02 23:37:54,871] ERROR [utils - 115:fail] - 10.70.42.147 - Change detection failed [root@dhcp43-48 ~]# [root@dhcp43-48 ~]# gluster v i ozone Volume Name: ozone Type: Tier Volume ID: 4611c8ba-4f32-409c-8858-81d55d2acc75 Status: Started Number of Bricks: 6 x 1 = 6 Transport-type: tcp Bricks: Brick1: 10.70.42.147:/rhs/thinbrick1/ozone/hdd Brick2: 10.70.43.48:/rhs/thinbrick1/ozone/hdd Brick3: 10.70.43.48:/rhs/thinbrick1/ozone/dd Brick4: 10.70.43.48:/rhs/thinbrick2/ozone/dd Brick5: 10.70.42.147:/rhs/thinbrick1/ozone/dd Brick6: 10.70.42.147:/rhs/thinbrick2/ozone/dd Options Reconfigured: storage.build-pgfid: on changelog.changelog: on geo-replication.ignore-pid-check: on geo-replication.indexing: on [root@dhcp43-48 ~]# [root@dhcp43-48 ~]# glusterfind session list usage: glusterfind [-h] {pre,create,list,post,delete} ... glusterfind: error: argument mode: invalid choice: 'session' (choose from 'pre', 'create', 'list', 'post', 'delete') [root@dhcp43-48 ~]# glusterfind list SESSION VOLUME SESSION TIME --------------------------------------------------------------------------- sess nash 2015-04-01 16:43:53 sess pluto 2015-04-02 16:46:13 sess ozone 2015-04-02 20:02:56 [root@dhcp43-48 ~]# [root@dhcp43-48 ~]# glusterfind pre sess ozone /tmp/out.txt --change-detector=changelog 10.70.43.48 - Change detection failed 10.70.43.48 - Change detection failed 10.70.43.48 - Change detection failed 10.70.42.147 - Change detection failed 10.70.42.147 - Change detection failed 10.70.42.147 - Change detection failed [root@dhcp43-48 ~]# [root@dhcp43-48 ~]# [root@dhcp43-48 ~]# --- Additional comment from Aravinda VK on 2015-04-02 10:21:35 EDT --- Issue is due to BZ 1207643. We don't need any patch for glusterfind to fix this apart from patches sent for 1207643 --- Additional comment from Aravinda VK on 2015-04-06 04:21:31 EDT --- Addition to the patch sent for BZ 1207643, we need to install "pyxattr" package. yum install pyxattr or pip install pyxattr --- Additional comment from Sweta Anandpara on 2015-04-06 08:14:25 EDT --- After installing the pyxattr package, the 'glusterfind pre' command still fails with the error: Change detection failed. However, I do see the changelog log files now. This is the error snippet: [2015-04-06 23:01:19,413] INFO [changelog - 286:changelog_crawl] - /rhs/brick2/dd Started Changelog Crawl. Start: 1427817769, End: 1428341464 [2015-04-06 23:01:19,481] INFO [changelog - 286:changelog_crawl] - /rhs/brick1/dd Started Changelog Crawl. Start: 1427817769, End: 1428341464 [2015-04-06 23:01:27,547] ERROR [utils - 115:fail] - Partial History available with Changelog [2015-04-06 23:15:23,152] INFO [changelog - 286:changelog_crawl] - /rhs/brick1/dd Started Changelog Crawl. Start: 1427817769, End: 1428342308 [2015-04-06 23:15:23,153] INFO [changelog - 286:changelog_crawl] - /rhs/brick2/dd Started Changelog Crawl. Start: 1427817769, End: 1428342308 [2015-04-06 23:15:29,187] ERROR [utils - 115:fail] - Partial History available with Changelog [2015-04-06 23:17:07,274] INFO [changelog - 286:changelog_crawl] - /rhs/brick2/dd Started Changelog Crawl. Start: 1427817769, End: 1428342412 [2015-04-06 23:17:07,295] INFO [changelog - 286:changelog_crawl] - /rhs/brick1/dd Started Changelog Crawl. Start: 1427817769, End: 1428342412 [2015-04-06 23:17:13,303] ERROR [utils - 115:fail] - Partial History available with Changelog The setup is 10.70.43.140 is in the same state, if you would want to have a look. --- Additional comment from Aravinda VK on 2015-04-07 07:26:53 EDT --- Changelog crash patch(http://review.gluster.org/10074) is merged and BZ 1209138 is opened to track package dependency. Moving this bug to MODIFIED.
Have not seen this issue again while running the feature functionality runs. The times when changelog has failed have been for failover/delete other issues which are being tracked by their corresponding bugs. Logs of the functionality run have been updated here: https://polarion.engineering.redhat.com/polarion/testrun-attachment/RHG3/glusterfs-3_7_1_3_RHEL6_7_FUSE/RHG3-5400_Logs_6.7_3.7.1-3_output_file_validation.odt Moving this to verified in 3.1 Everglades
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHSA-2015-1495.html