Unit in "agent-state: error" cannot be resolved
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
juju-core |
Fix Released
|
Medium
|
Unassigned |
Bug Description
I have a unit which is in "agent-state: error":
$ juju status
services:
charmname:
units:
charmname/0:
machine: "2"
charmname/1:
machine: "15"
but cannot be resolved:
$ juju resolved --retry charmname/1
error: cannot set resolved mode for unit "charmname/1": already resolved
The symptoms sound similar to LP: #1175031, but in a much newer version. Through a combination of "juju upgrade-charm --force charmname", "juju resolved charmname/1" and restarting the charm agent on the unit, I've been occasionally able to get it to retry, but no matter the outcome of the hooks, it doesn't clear the error state.
=== Agent start example ===
2014-01-16 20:50:27 INFO juju.jujud unit.go:72 unit agent unit-charmname-1 start
2014-01-16 20:50:27 INFO juju runner.go:253 worker: start "state"
2014-01-16 20:50:27 INFO juju.state open.go:68 opening state; mongo addresses: ["10.33.
2014-01-16 20:50:27 INFO juju runner.go:253 worker: start "api"
2014-01-16 20:50:27 INFO juju apiclient.go:106 state/api: dialing "wss://
2014-01-16 20:50:27 INFO juju apiclient.go:116 state/api: connection established
2014-01-16 20:50:27 INFO juju.state open.go:106 connection established
2014-01-16 20:50:28 INFO juju runner.go:253 worker: start "upgrader"
2014-01-16 20:50:28 INFO juju runner.go:253 worker: start "uniter"
2014-01-16 20:50:28 INFO juju.worker.uniter uniter.go:83 unit "charmname/1" started
2014-01-16 20:50:28 INFO juju.worker.
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:408 ModeInit starting
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:29 updating unit addresses
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:52 reconciling relation state
2014-01-16 20:50:28 INFO juju.worker.uniter uniter.go:508 joining relation "subcharm:a charmname:a"
2014-01-16 20:50:28 INFO juju.worker.uniter uniter.go:530 joined relation "subcharm:a charmname:a"
2014-01-16 20:50:28 DEBUG juju.worker.uniter modes.go:410 ModeInit exiting
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:408 ModeContinue starting
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:65 loading uniter state
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:109 awaiting error resolution for "upgrade-charm" hook
2014-01-16 20:50:28 DEBUG juju.worker.uniter modes.go:410 ModeContinue exiting
2014-01-16 20:50:28 INFO juju.worker.uniter modes.go:408 ModeHookError starting
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
2014-01-16 20:50:28 DEBUG juju.worker.
=== Agent start example w/ upgrade-charm retry ===
2014-01-16 20:30:38 INFO juju.jujud unit.go:72 unit agent unit-charmname-1 start
2014-01-16 20:30:38 INFO juju runner.go:253 worker: start "state"
2014-01-16 20:30:38 INFO juju.state open.go:68 opening state; mongo addresses: ["10.33.
2014-01-16 20:30:38 INFO juju runner.go:253 worker: start "api"
2014-01-16 20:30:38 INFO juju apiclient.go:106 state/api: dialing "wss://
2014-01-16 20:30:38 INFO juju apiclient.go:116 state/api: connection established
2014-01-16 20:30:38 INFO juju.state open.go:106 connection established
2014-01-16 20:30:39 INFO juju runner.go:253 worker: start "uniter"
2014-01-16 20:30:39 INFO juju.worker.uniter uniter.go:83 unit "charmname/1" started
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:408 ModeInit starting
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:29 updating unit addresses
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 INFO juju runner.go:253 worker: start "upgrader"
2014-01-16 20:30:39 INFO juju.worker.
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:52 reconciling relation state
2014-01-16 20:30:39 INFO juju.worker.uniter uniter.go:508 joining relation "subcharm:a charmname:a"
2014-01-16 20:30:39 INFO juju.worker.uniter uniter.go:530 joined relation "subcharm:a charmname:a"
2014-01-16 20:30:39 DEBUG juju.worker.uniter modes.go:410 ModeInit exiting
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:408 ModeContinue starting
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:65 loading uniter state
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:109 awaiting error resolution for "upgrade-charm" hook
2014-01-16 20:30:39 DEBUG juju.worker.uniter modes.go:410 ModeContinue exiting
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:408 ModeHookError starting
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.uniter modes.go:410 ModeHookError exiting
2014-01-16 20:30:39 INFO juju.worker.uniter modes.go:408 ModeUpgrading local:precise/
2014-01-16 20:30:39 INFO juju.worker.uniter uniter.go:230 fetching charm "local:
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 INFO juju charm.go:56 worker/
2014-01-16 20:30:39 INFO juju charm.go:68 worker/
2014-01-16 20:30:39 INFO juju charm.go:80 worker/
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 INFO juju.worker.uniter uniter.go:253 deploying charm "local:
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 DEBUG juju.worker.
2014-01-16 20:30:39 INFO juju deployer.go:144 worker/
2014-01-16 20:30:39 INFO juju deployer.go:164 worker/
2014-01-16 20:30:40 INFO juju deployer.go:98 worker/
2014-01-16 20:30:40 INFO juju.worker.uniter uniter.go:264 charm "local:
2014-01-16 20:30:40 DEBUG juju.worker.uniter modes.go:410 ModeUpgrading local:precise/
2014-01-16 20:30:40 INFO juju.worker.uniter modes.go:408 ModeContinue starting
2014-01-16 20:30:40 INFO juju.worker.uniter modes.go:109 awaiting error resolution for "upgrade-charm" hook
2014-01-16 20:30:40 DEBUG juju.worker.uniter modes.go:410 ModeContinue exiting
2014-01-16 20:30:40 INFO juju.worker.uniter modes.go:408 ModeHookError starting
2014-01-16 20:30:40 DEBUG juju.worker.
2014-01-16 20:30:40 DEBUG juju.worker.
2014-01-16 20:30:40 DEBUG juju.worker.
2014-01-16 20:30:40 DEBUG juju.worker.
2014-01-16 20:30:43 DEBUG juju.worker.
tags: | added: resolved |
tags: | added: 1.14.1 |
Changed in juju-core: | |
status: | New → Triaged |
importance: | Undecided → High |
tags: | added: upgrade-charm |
Changed in juju-core: | |
importance: | High → Medium |
tags: | added: canonical-bootstack |
tags: | added: canonical-is |
I have a similar issue:
$ juju status rabbitmq-server/1 | grep -A 3 rabbitmq-server/1: server/ 1:
agent- state: error
agent- state-info: 'hook failed: "config-changed"'
agent- version: 1.20.9.1
rabbitmq-
$ juju resolved rabbitmq-server/1 server/ 1": already resolved
ERROR cannot set resolved mode for unit "rabbitmq-
When restarting jujud I can see it retrying the config-changed hook w/o issue, but the error state remains:
# restart jujud-unit- rabbitmq- server- 1
-- log excerpt of unit-rabbitmq- server- 1
2015-02-05 13:42:42 INFO juju.worker.uniter uniter.go:115 unit "rabbitmq-server/1" started uniter. filter filter.go:507 charm check skipped, not yet installed. uniter. filter filter.go:325 got service change uniter. filter filter.go:521 no new charm event uniter. filter filter.go:317 got unit change uniter. filter filter.go:341 got relations change uniter. filter filter.go:333 got config change uniter. filter filter.go:337 preparing new config event uniter. filter filter.go:440 discarded config event server/ 1:config- changed: 543336471173697 9567"; dir "/var/lib/ juju/agents/ unit-rabbitmq- server- 1/charm" server/ 1:config- changed: 543336471173697 9567"; dir "/var/lib/ juju/agents/ unit-rabbitmq- server- 1/charm" server/ 1:config- changed: 543336471173697 9567"; dir "/var/lib/ juju/agents/ unit-rabbitmq- server- 1/charm" security. ubuntu. com trusty-security InRelease
2015-02-05 13:42:42 DEBUG juju.worker.uniter runlistener.go:84 juju-run listener running
2015-02-05 13:42:42 INFO juju.worker.uniter modes.go:396 ModeContinue starting
2015-02-05 13:42:42 INFO juju.worker.uniter modes.go:31 loading uniter state
2015-02-05 13:42:42 INFO juju.worker.uniter modes.go:48 continuing after "config-changed" hook
2015-02-05 13:42:42 DEBUG juju.worker.uniter modes.go:398 ModeContinue exiting
2015-02-05 13:42:42 INFO juju.worker.uniter modes.go:396 ModeConfigChanged starting
2015-02-05 13:42:42 DEBUG juju.worker.uniter uniter.go:757 new environment change
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 DEBUG juju.worker.
2015-02-05 13:42:42 INFO juju.worker.uniter uniter.go:480 running "config-changed" hook
2015-02-05 13:42:43 INFO worker.uniter.jujuc server.go:102 running hook tool "config-get" ["source" "--format=json"]
2015-02-05 13:42:43 DEBUG worker.uniter.jujuc server.go:103 hook context id "rabbitmq-
2015-02-05 13:42:43 INFO worker.uniter.jujuc server.go:102 running hook tool "config-get" ["key" "--format=json"]
2015-02-05 13:42:43 DEBUG worker.uniter.jujuc server.go:103 hook context id "rabbitmq-
2015-02-05 13:42:43 INFO worker.uniter.jujuc server.go:102 running hook tool "juju-log" ["Source is not present. Skipping"]
2015-02-05 13:42:43 DEBUG worker.uniter.jujuc server.go:103 hook context id "rabbitmq-
2015-02-05 13:42:43 INFO juju-log Source is not present. Skipping
2015-02-05 13:42:43 DEBUG juju.worker.rsyslog worker.go:164 Reloading rsyslog configuration
2015-02-05 13:42:43 INFO config-changed Ign http://
2015-02-05 13:...