wedged unit state error but can't be resolved
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
juju-core |
Fix Released
|
Critical
|
William Reade |
Bug Description
$ juju-status snippet
cache-lb:
charm: local:precise/
exposed: false
relations:
reverseproxy:
- wordpress
units:
cache-lb/0:
machine: "6"
$ juju resolved cache-lb/0
error: cannot set resolved mode for unit "cache-lb/0": already resolved
I was trying some automated resolved retry logic in deployer as can be seen in the unit log, but after a few retries it gets wedged.
http://
incidentally the rel hook error is due to the lack of -r support (reported separately).
Related branches
- Juju Engineering: Pending requested
-
Diff: 207 lines (+100/-17)4 files modifiedworker/uniter/filter.go (+51/-5)
worker/uniter/filter_test.go (+20/-10)
worker/uniter/modes.go (+2/-2)
worker/uniter/uniter_test.go (+27/-0)
Changed in juju-core: | |
assignee: | nobody → William Reade (fwereade) |
Changed in juju-core: | |
status: | Triaged → In Progress |
Changed in juju-core: | |
status: | In Progress → Fix Committed |
Changed in juju-core: | |
milestone: | none → 1.11.1 |
Changed in juju-core: | |
status: | Fix Committed → Fix Released |
Next time you encounter this, please report back what happens if you ssh to the machine and bounce the unit agent process (something like `sudo restart jujud-unit- cache-lb- 0`). If this resolves the issue, it should eliminate the possibility of serious state brokenness and isolate it to the unit agent itself.