upgrade 1.14.1 to 1.16.2 leaves machine agents not upgraded
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
juju-core |
Won't Fix
|
High
|
Unassigned |
Bug Description
Running on EC2 with a small instance.
Ran the following command this morning: juju upgrade-juju
No errors reported on console after running
All the nodes are down, but services are otherwise running normally.
I would expect all services to restart with the new 1.16.2 symlinked agent and status to be started after a small time.
`juju status` http://
machine-0 info:
machine-0.log from that time: http://
machine agent is linked to 1.14.1 tools, but has 1.16.2 downloaded
```
ubuntu@
total 16
drwxr-xr-x 4 root root 4096 Nov 4 22:06 .
drwxr-xr-x 5 root root 4096 Sep 24 15:18 ..
drwxr-xr-x 2 root root 4096 Sep 24 15:17 1.14.1-
drwx------ 2 root root 4096 Nov 4 22:16 1.16.2-
lrwxrwxrwx 1 root root 20 Sep 24 15:19 machine-0 -> 1.14.1-
```
Machine-2 information:
upstart job for agent: http://
log from time of upgrade-juju: http://
unit on the machine is using the 1.16.2 tools, the machine agent is using the old 1.14.1 tools
```
ubuntu@
total 16
drwxr-xr-x 4 root root 4096 Nov 4 22:04 .
drwxr-xr-x 5 root root 4096 Sep 24 16:14 ..
drwxr-xr-x 2 root root 4096 Sep 24 16:14 1.14.1-
drwx------ 2 root root 4096 Nov 4 22:04 1.16.2-
lrwxrwxrwx 1 root root 20 Sep 24 16:14 machine-2 -> 1.14.1-
lrwxrwxrwx 1 root root 20 Nov 4 22:04 unit-picatic-api-0 -> 1.16.2-
```
tags: | added: juju-gui |
tags: |
added: upgrade-juju removed: juju-gui |
Changed in juju-core: | |
importance: | Undecided → High |
status: | New → Triaged |
Changed in juju-core: | |
status: | Triaged → Won't Fix |
Would appreciate any thoughts on how to bring these node agents back online. This is currently sitting on our production cluster, cloning and redeploying is an option, but would like to see if we can get this one rolling again.