upgrade-juju from 2.0.0 to 2.0.1 failed

Bug #1638686 reported by George Kraft
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Canonical Juju
New
Undecided
Unassigned

Bug Description

Using 2.0.1 client on MAAS.

Originally bootstrapped on Juju 2.0-rc2 and have upgraded a number of times via `juju upgrade-juju`.

When we tried to upgrade from 2.0.0 to 2.0.1 with `juju upgrade-juju -m controller`, the command seemed to be successful; however, the controller agent does not appear to have been upgraded.

On the controller machine:

$ sudo cat /var/log/juju/machine-0.log | grep -i "running jujud"
2016-10-03 15:23:48 INFO juju.cmd supercommand.go:63 running jujud [2.0-rc2 gc go1.6.2]
2016-10-12 20:42:58 INFO juju.cmd supercommand.go:63 running jujud [2.0-rc3.1 gc go1.6.2]
2016-10-14 16:41:13 INFO juju.cmd supercommand.go:63 running jujud [2.0.0.1 gc go1.6.2]

It appears the 2.0.1 tools were not copied in:

$ ls -1 /var/lib/juju/tools
2.0.0.1-xenial-amd64
2.0-rc2-xenial-amd64
2.0-rc3.1-xenial-amd64
machine-0

If we try to run the upgrade-juju command again, we get an error:

$ juju upgrade-juju -m controller
ERROR unexpectedly bad version "2.0.1.1-zesty-amd64" in tools storage: unknown OS for series: "zesty"

This message is being spammed in the attached controller logs as well. The controller is running xenial.

Tags: v-pil
Revision history for this message
George Kraft (cynerva) wrote :
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.