-
Committer:
Scott Moser
-
Date:
2012-01-05 18:56:54 UTC
-
Revision ID:
smoser@ubuntu.com-20120105185654-1pun1mhgdcdopju9
increase the wait time for a volume to go to 'available'
This increases the time after a 'detach-vol' command is made before it must be
in 'available' state dramatically. Previously we were issuing the command and
then checking 5 times with 3 second sleeps. Now, check 20 times with 5 second
sleeps.
There have been I think 3 builds in the past week that have failed because we
didn't wait long enough for detach'd volume to become available.
The one below is from oneiric desktop 20120105.
It shows that after 47 seconds the volume was not yet available, but then
within 63 seconds it was.
sa-east-1:copied data to /dev/sdi
sa-east-1: Thu Jan 5 17:09:13 UTC 2012: detaching vol-1fa3d577
sa-east-1: Thu Jan 5 17:09:40 UTC 2012: after detach try 1/5, state=in-use
sa-east-1: Thu Jan 5 17:09:45 UTC 2012: after detach try 2/5, state=in-use
sa-east-1: Thu Jan 5 17:09:49 UTC 2012: after detach try 3/5, state=in-use
sa-east-1: Thu Jan 5 17:09:56 UTC 2012: after detach try 4/5, state=in-use
sa-east-1: Thu Jan 5 17:10:00 UTC 2012: after detach try 5/5, state=in-use
sa-east-1: after 5 waits, vol-1fa3d577, state=in-use (not 'available')
sa-east-1: Thu Jan 5 17:10:03 UTC 2012: cmd failed [1/5]: detach_vol vol-1fa3d577 /tmp/ec2-image2ebs.Q18bND/vol_detach.info
sa-east-1: Thu Jan 5 17:10:18 UTC 2012: detaching vol-1fa3d577
Client.IncorrectState: Volume 'vol-1fa3d577'is in the 'available' state.
sa-east-1: failed to detach vol-1fa3d577
sa-east-1: Thu Jan 5 17:10:29 UTC 2012: cmd failed [2/5]: detach_vol vol-1fa3d577 /tmp/ec2-image2ebs.Q18bND/vol_detach.info
sa-east-1: Thu Jan 5 17:10:44 UTC 2012: detaching vol-1fa3d577
Client.IncorrectState: Volume 'vol-1fa3d577'is in the 'available' state.