Hello Community, I had a power issue last night and all the systems went down. When they powered up the arbiter was offline (network issue) which was resolved and heal passed for almost all files but 2 per volume. The strange thing is that the arbiter's copy is older and the 'trusted.gfid' is different on the arbiter. In which cases this can happen ?
I guess the heal is not passing ,as the 'trusted.gfid' is different while 'trusted.gfid2path.<long_string>' is the same. Here is the output of getfattr and I need to realise if this is a possible situation or it is a bug that needs reporting: Brick1:[root@ovirt1 ~]# getfattr -d -m . -e hex /gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673b b2.meta getfattr: Removing leading '/' from absolute path names # file: gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673bb2.meta security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000 trusted.afr.data_fast-client-2=0x000000020000000200000000 trusted.afr.dirty=0x000000000000000000000000 trusted.gfid=0x2408e73df137431e99c663783b8bb356 trusted.gfid2path.27d407743ed6191b=0x35333430653763322d666333302d343461382d613633392d3762613138636138653033362f35373161356562342d656239322d343739632d383330352d3739656161303637336262322e6d65746 1 trusted.glusterfs.mdata=0x010000000000000000000000005dcc1eae0000000023c985d9000000005dcc1eae000000001f4c18bf000000005dcc1eae000000001ed22d12 trusted.glusterfs.shard.block-size=0x0000000004000000 trusted.glusterfs.shard.file-size=0x00000000000001b4000000000000000000000000000000010000000000000000 Brick2:[root@ovirt2 ~]# getfattr -d -m . -e hex /gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673b b2.meta getfattr: Removing leading '/' from absolute path names # file: gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673bb2.meta security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000 trusted.afr.data_fast-client-2=0x000000020000000200000000 trusted.afr.dirty=0x000000000000000000000000 trusted.gfid=0x2408e73df137431e99c663783b8bb356 trusted.gfid2path.27d407743ed6191b=0x35333430653763322d666333302d343461382d613633392d3762613138636138653033362f35373161356562342d656239322d343739632d383330352d3739656161303637336262322e6d65746 1 trusted.glusterfs.mdata=0x010000000000000000000000005dcc1eae0000000023c985d9000000005dcc1eae000000001f4c18bf000000005dcc1eae000000001ed22d12 trusted.glusterfs.shard.block-size=0x0000000004000000 trusted.glusterfs.shard.file-size=0x00000000000001b4000000000000000000000000000000010000000000000000 Arbiter:[root@ovirt3 ~]# getfattr -d -m . -e hex /gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673b b2.meta getfattr: Removing leading '/' from absolute path names # file: gluster_bricks/data_fast/data_fast/396604d9-2a9e-49cd-9563-fdc79981f67b/images/25107759-f56b-4d94-872e-0f9b145886d1/571a5eb4-eb92-479c-8305-79eaa0673bb2.meta security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000 trusted.afr.dirty=0x000000000000000000000000 trusted.gfid=0x48c9c4ebe75c41f7ac45ea4943e9a65a trusted.gfid2path.27d407743ed6191b=0x35333430653763322d666333302d343461382d613633392d3762613138636138653033362f35373161356562342d656239322d343739632d383330352d3739656161303637336262322e6d65746 1 trusted.glusterfs.mdata=0x010000000000000000000000005dcb3a520000000011c7ae2c000000005dcb3a520000000011690f7b000000005dcb3a520000000011690f7b trusted.glusterfs.shard.block-size=0x0000000004000000 trusted.glusterfs.shard.file-size=0x00000000000001b4000000000000000000000000000000010000000000000000 Also, as per 'Troubleshooting Split-Brains - Gluster Docs' there should be another 2 copies of the changelog (local + remote bricks) but in my case they are missing. Why is it like that ? "Each file in a brick maintains the changelog of itself and that of the files present in all the other bricks in its replica set as seen by that brick." | | | | Troubleshooting Split-Brains - Gluster Docs Project documentation for Gluster Filesystem | | | Thanks in advance for your response. Best Regards,Strahil Nikolov
________ Community Meeting Calendar: APAC Schedule - Every 2nd and 4th Tuesday at 11:30 AM IST Bridge: https://bluejeans.com/118564314 NA/EMEA Schedule - Every 1st and 3rd Tuesday at 01:00 PM EDT Bridge: https://bluejeans.com/118564314 Gluster-users mailing list Gluster-users@gluster.org https://lists.gluster.org/mailman/listinfo/gluster-users
