Description of problem: ----------------------- (Thanks Ravi for the catch!) Create rep2/3 volume on 3.3 bits.. I see that client-io-threads is getting loaded for newly created volumes : [root@gqas013 rep]# cat rep.tcp-fuse.vol volume rep-client-0 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks10/A1 option remote-host gqas013 option ping-timeout 42 end-volume volume rep-client-1 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks10/A1 option remote-host gqas005 option ping-timeout 42 end-volume volume rep-client-2 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks10/A1 option remote-host gqas006 option ping-timeout 42 end-volume volume rep-client-3 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks11/A1 option remote-host gqas013 option ping-timeout 42 end-volume volume rep-client-4 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks11/A1 option remote-host gqas005 option ping-timeout 42 end-volume volume rep-client-5 type protocol/client option send-gids true option event-threads 4 option transport.address-family inet option transport-type tcp option remote-subvolume /bricks11/A1 option remote-host gqas008 option ping-timeout 42 end-volume volume rep-replicate-0 type cluster/replicate option use-compound-fops off subvolumes rep-client-0 rep-client-1 rep-client-2 end-volume volume rep-replicate-1 type cluster/replicate option use-compound-fops off subvolumes rep-client-3 rep-client-4 rep-client-5 end-volume volume rep-dht type cluster/distribute option lock-migration off option lookup-optimize on subvolumes rep-replicate-0 rep-replicate-1 end-volume volume rep-write-behind type performance/write-behind subvolumes rep-dht end-volume volume rep-read-ahead type performance/read-ahead subvolumes rep-write-behind end-volume volume rep-readdir-ahead type performance/readdir-ahead option rda-cache-limit 10MB option rda-request-size 131072 option parallel-readdir off subvolumes rep-read-ahead end-volume volume rep-io-cache type performance/io-cache subvolumes rep-readdir-ahead end-volume volume rep-quick-read type performance/quick-read subvolumes rep-io-cache end-volume volume rep-open-behind type performance/open-behind subvolumes rep-quick-read end-volume volume rep-md-cache type performance/md-cache subvolumes rep-open-behind end-volume volume rep-io-threads type performance/io-threads // THIS HERE subvolumes rep-md-cache end-volume volume rep type debug/io-stats option count-fop-hits off option latency-measurement off option log-level INFO subvolumes rep-io-threads end-volume [root@gqas013 rep]# Version-Release number of selected component (if applicable): ------------------------------------------------------------- 3.8.4-25 How reproducible: ----------------- Always. Additional info: ----------------- [root@gqas013 rep]# gluster v info rep Volume Name: rep Type: Distributed-Replicate Volume ID: ef19f98e-18cd-417a-bada-769afafd3125 Status: Started Snapshot Count: 0 Number of Bricks: 2 x 3 = 6 Transport-type: tcp Bricks: Brick1: gqas013:/bricks10/A1 Brick2: gqas005:/bricks10/A1 Brick3: gqas006:/bricks10/A1 Brick4: gqas013:/bricks11/A1 Brick5: gqas005:/bricks11/A1 Brick6: gqas008:/bricks11/A1 Options Reconfigured: ganesha.enable: on features.cache-invalidation: off client.event-threads: 4 server.event-threads: 4 cluster.lookup-optimize: on transport.address-family: inet nfs.disable: on nfs-ganesha: enable cluster.enable-shared-storage: enable [root@gqas013 rep]#
This will be a downstream only patch. Details are on the commit message why this is a regression: https://code.engineering.redhat.com/gerrit/107558
Verified on 3.8.4-27 on newly created replicate and dist rep volumes. [root@gqas013 distrep]# cat distrep.tcp-fuse.vol |grep -i threads [root@gqas013 distrep]# [root@gqas013 distrep]# cat ../testvol/testvol.tcp-fuse.vol |grep -i threads [root@gqas013 distrep]# [root@gqas013 distrep]# [root@gqas013 distrep]# gluster v info Volume Name: distrep Type: Distributed-Replicate Volume ID: 65d3ac82-7c47-4631-bbd3-509cc7e84f7b Status: Started Snapshot Count: 0 Number of Bricks: 2 x 2 = 4 Transport-type: tcp Bricks: Brick1: gqas013:/bricks2/A11 Brick2: gqas005:/bricks2/A22 Brick3: gqas013:/bricks3/A11 Brick4: gqas005:/bricks4/A12 Options Reconfigured: transport.address-family: inet nfs.disable: on Volume Name: testvol Type: Replicate Volume ID: 76b11dfe-7f9f-4b3a-a0f0-1814cf936510 Status: Started Snapshot Count: 0 Number of Bricks: 1 x 2 = 2 Transport-type: tcp Bricks: Brick1: gqas013:/bricks1/A2 Brick2: gqas005:/bricks1/A3 Options Reconfigured: transport.address-family: inet nfs.disable: on [root@gqas013 distrep]#
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://access.redhat.com/errata/RHBA-2017:2774