Re: [Gluster-devel] Spurious failure - ./tests/bugs/bug-859581.t
On 06/18/2014 10:13 AM, Pranith Kumar Karampuri wrote: On 06/18/2014 10:11 AM, Atin Mukherjee wrote: On 06/18/2014 10:04 AM, Pranith Kumar Karampuri wrote: On 06/18/2014 09:39 AM, Atin Mukherjee wrote: Pranith, Regression test mentioned in $SUBJECT failed (testcase : 14 16) Console log can be found at http://build.gluster.org/job/rackspace-regression-2GB/227/consoleFull My initial suspect is on HEAL_TIMEOUT (set to 60 seconds) where healing might not have been completed within this time frame and i.e. why EXPECT_WITHIN fails. I am not sure on what basis this HEAL_TIMEOUT's value was derived. Probably you would be the better to analyse it. Having a larger time out value might help here? I don't think it is a spurious failure. There seems to be a bug in afr-v2. I will have to fix that. If its not a spurious failure why its not failing every time? Depends on which subvolume afr picks in readdir. If it reads the one with the directory it will succeed. Otherwise it will fail. The bug is easy to hit of the steps are carried out in a 2 node setup. I have created a BZ https://bugzilla.redhat.com/show_bug.cgi?id=1112158 and am looking into it. Pranith Pranith Cheers, Atin ___ Gluster-devel mailing list Gluster-devel@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-devel ___ Gluster-devel mailing list Gluster-devel@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-devel
Re: [Gluster-devel] Spurious failure - ./tests/bugs/bug-859581.t
On 06/18/2014 10:04 AM, Pranith Kumar Karampuri wrote: On 06/18/2014 09:39 AM, Atin Mukherjee wrote: Pranith, Regression test mentioned in $SUBJECT failed (testcase : 14 16) Console log can be found at http://build.gluster.org/job/rackspace-regression-2GB/227/consoleFull My initial suspect is on HEAL_TIMEOUT (set to 60 seconds) where healing might not have been completed within this time frame and i.e. why EXPECT_WITHIN fails. I am not sure on what basis this HEAL_TIMEOUT's value was derived. Probably you would be the better to analyse it. Having a larger time out value might help here? I don't think it is a spurious failure. There seems to be a bug in afr-v2. I will have to fix that. If its not a spurious failure why its not failing every time? Pranith Cheers, Atin ___ Gluster-devel mailing list Gluster-devel@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-devel
Re: [Gluster-devel] Spurious failure - ./tests/bugs/bug-859581.t
On 06/18/2014 10:11 AM, Atin Mukherjee wrote: On 06/18/2014 10:04 AM, Pranith Kumar Karampuri wrote: On 06/18/2014 09:39 AM, Atin Mukherjee wrote: Pranith, Regression test mentioned in $SUBJECT failed (testcase : 14 16) Console log can be found at http://build.gluster.org/job/rackspace-regression-2GB/227/consoleFull My initial suspect is on HEAL_TIMEOUT (set to 60 seconds) where healing might not have been completed within this time frame and i.e. why EXPECT_WITHIN fails. I am not sure on what basis this HEAL_TIMEOUT's value was derived. Probably you would be the better to analyse it. Having a larger time out value might help here? I don't think it is a spurious failure. There seems to be a bug in afr-v2. I will have to fix that. If its not a spurious failure why its not failing every time? Depends on which subvolume afr picks in readdir. If it reads the one with the directory it will succeed. Otherwise it will fail. Pranith Pranith Cheers, Atin ___ Gluster-devel mailing list Gluster-devel@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-devel