Sachin and I looked at the failure. Current guess is that glusterd_2 had not yet completed the handshake with glusterd_1 and hence did not know about the option set. KP suggested that instead of having a sleep before this command, we could get peer status and verify that it is 1 and then get the vol info. Although even this does not make the test fully deterministic, we will be closer to it. Sachin will send out a patch for the same. Raghavendra Talur ----- Original Message ----- From: "Pranith Kumar Karampuri" <pkarampu@xxxxxxxxxx> To: "Sachin Pandit" <spandit@xxxxxxxxxx> Cc: "Gluster Devel" <gluster-devel@xxxxxxxxxxx> Sent: Thursday, June 12, 2014 9:54:03 AM Subject: Re: spurious regression failure in tests/bugs/bug-1104642.t Check the logs to find the reason. Pranith. On 06/12/2014 09:24 AM, Sachin Pandit wrote: > I am not hitting this even after running the test case in a loop. > I'll update in this thread once I find out the root cause of the failure. > > ~ Sachin > > ----- Original Message ----- > From: "Sachin Pandit" <spandit@xxxxxxxxxx> > To: "Pranith Kumar Karampuri" <pkarampu@xxxxxxxxxx> > Cc: "Gluster Devel" <gluster-devel@xxxxxxxxxxx> > Sent: Thursday, June 12, 2014 8:50:40 AM > Subject: Re: spurious regression failure in tests/bugs/bug-1104642.t > > I will look into this. > > ----- Original Message ----- > From: "Pranith Kumar Karampuri" <pkarampu@xxxxxxxxxx> > To: "Gluster Devel" <gluster-devel@xxxxxxxxxxx> > Cc: rtalur@xxxxxxxxxx, spandit@xxxxxxxxxx > Sent: Wednesday, June 11, 2014 9:08:44 PM > Subject: spurious regression failure in tests/bugs/bug-1104642.t > > Raghavendra/Sachin, > Could one of you guys take a look at this please. > > pk1@localhost - ~/workspace/gerrit-repo (master) > 21:04:46 :) ⚡ ~/.scripts/regression.py > http://build.gluster.org/job/regression/4831/consoleFull > Patch ==> http://review.gluster.com/#/c/7994/2 > Author ==> Raghavendra Talur rtalur@xxxxxxxxxx > Build triggered by ==> amarts > Build-url ==> http://build.gluster.org/job/regression/4831/consoleFull > Download-log-at ==> > http://build.gluster.org:443/logs/regression/glusterfs-logs-20140611:08:39:04.tgz > Test written by ==> Author: Sachin Pandit <spandit@xxxxxxxxxx> > > ./tests/bugs/bug-1104642.t [13] > 0 #!/bin/bash > 1 > 2 . $(dirname $0)/../include.rc > 3 . $(dirname $0)/../volume.rc > 4 . $(dirname $0)/../cluster.rc > 5 > 6 > 7 function get_value() > 8 { > 9 local key=$1 > 10 local var="CLI_$2" > 11 > 12 eval cli_index=\$$var > 13 > 14 $cli_index volume info | grep "^$key"\ > 15 | sed 's/.*: //' > 16 } > 17 > 18 cleanup > 19 > 20 TEST launch_cluster 2 > 21 > 22 TEST $CLI_1 peer probe $H2; > 23 EXPECT_WITHIN $PROBE_TIMEOUT 1 peer_count > 24 > 25 TEST $CLI_1 volume create $V0 $H1:$B1/${V0}0 $H2:$B2/${V0}1 > 26 EXPECT "$V0" get_value 'Volume Name' 1 > 27 EXPECT "Created" get_value 'Status' 1 > 28 > 29 TEST $CLI_1 volume start $V0 > 30 EXPECT "Started" get_value 'Status' 1 > 31 > 32 #Bring down 2nd glusterd > 33 TEST kill_glusterd 2 > 34 > 35 #set the volume all options from the 1st glusterd > 36 TEST $CLI_1 volume set all cluster.server-quorum-ratio 80 > 37 > 38 #Bring back the 2nd glusterd > 39 TEST $glusterd_2 > 40 > 41 #Verify whether the value has been synced > 42 EXPECT '80' get_value 'cluster.server-quorum-ratio' 1 > ***43 EXPECT '80' get_value 'cluster.server-quorum-ratio' 2 > 44 > 45 cleanup; > > Pranith > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxxx > http://supercolony.gluster.org/mailman/listinfo/gluster-devel _______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx http://supercolony.gluster.org/mailman/listinfo/gluster-devel -- Thanks! Raghavendra Talur | Red Hat Storage Developer | Bangalore | +918039245176 _______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx http://supercolony.gluster.org/mailman/listinfo/gluster-devel
