gce Invalid volume ID destroying environment
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
juju-core |
Invalid
|
High
|
Unassigned | ||
1.25 |
Fix Released
|
High
|
Anastasia |
Bug Description
While destroying GCE environment, Juju gets a "malformed volume id" error when other environments are also running
Juju version used 1.25.0 to 1,25.4 and master
Suspicious code:
https:/
It seems the above code is trying to split the volume id using "--" as a separator but the id doesn't contain "--". Are we passing incorrect volume id?
Log:
http://
2016-03-07 19:00:07 ERROR juju.cmd supercommand.go:429 environment destruction failed: destroying storage: destroying volumes: invalid volume id "juju-1ee04202-
How to reproduce:
Bootstrap two gce controllers or state-servers.
Destroy one controller or state-server
The error is reported because of the other disk for other server.
Juju exists with an error
description: | updated |
Changed in juju-core: | |
status: | New → Triaged |
tags: | added: regression |
description: | updated |
Changed in juju-core: | |
status: | Triaged → Invalid |
The first appearance of the error in 1.25 is reports. vapour. ws/releases/ 3397 1.25.2 at /github. com/juju/ juju/commit/ e13b531a147d162 efba21c2ed707b5 3030da5ebe
http://
https:/
The last revision tested without this error is reports. vapour. ws/releases/ 3391 1.25.2 /github. com/juju/ juju/commit/ 38594d27eddcf0b 54d15eacbaafd92 85a8f533f8
http://
https:/
The suspect commits are /github. com/juju/ juju/commit/ 68d2677e200c5d7 180650231684a0d fd2b2c38a3 /github. com/juju/ juju/commit/ e13b531a147d162 efba21c2ed707b5 3030da5ebe
https:/
https:/
As both commits deal with upgrade. I suspect the cloud changed