2.1.2 juju status can't parse response from 2.2-alpha1 agent
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Canonical Juju |
Expired
|
Medium
|
Unassigned |
Bug Description
I have a model with agents running 2.2-alpha1 on top of a landscape autopilot openstack cloud. A 'juju status' with 2.1.2 results in an error, but works with 2.2-beta1. The error without --debug is:
ERROR codec.ReadHeader error: error receiving message: unexpected end of JSON input
This failure only occurs with one of the three models on this controller. All models are running 2.2-alpha1. The model that fails contains some machines which failed to deploy due to an issue with the underlying openstack cloud and are stuck in an error state. The 2.2-beta1 version of juju status displays a lot of additional error info for these instances. The full log of juju status from 2.2-beta1 is attached.
Expected behavior:
I would expect any 2.x series "juju status" to display status regardless of the agent-version of the model.
Here's what I see:
[While using juju 2.1.2]
$ juju model-config|grep agent-version
agent-version model 2.2-alpha1
$ juju status --debug
13:39:41 INFO juju.cmd supercommand.go:63 running juju [2.1.2 gc go1.6.2]
13:39:41 DEBUG juju.cmd supercommand.go:64 args: []string{"juju", "status", "--debug"}
13:39:41 INFO juju.juju api.go:72 connecting to API addresses: [10.96.127.8:17070 10.100.0.11:17070]
13:39:41 DEBUG juju.api apiclient.go:590 dialing "wss://
13:39:41 DEBUG juju.api apiclient.go:590 dialing "wss://
13:39:41 DEBUG juju.api apiclient.go:606 will retry after error dialing websocket: websocket.Dial wss://10.
13:39:43 DEBUG juju.api apiclient.go:593 successfully dialed "wss://
13:39:43 INFO juju.api apiclient.go:521 connection established to "wss://
13:39:43 DEBUG juju.juju api.go:263 API hostnames unchanged - not resolving
13:39:43 DEBUG juju.api monitor.go:35 RPC connection died
13:39:43 ERROR cmd supercommand.go:458 codec.ReadHeader error: error receiving message: unexpected end of JSON input
13:39:43 DEBUG cmd supercommand.go:459 (error details: [{github.
[While using juju 2.2-beta1]
$ juju version
2.2-beta1-
$ juju model-config|grep agent-version
agent-version model 2.2-alpha1
$ juju status
Model Controller Cloud/Region Version
long-running lrc-region1 lrc/region1 2.2-alpha1
App Version Status Scale Charm Store Rev OS Notes
landscape-
landscape-
trusty-client waiting 0/1 landscape-devenv local 0 ubuntu
trusty-server waiting 0/2 landscape-devenv local 1 ubuntu
ubuntu 16.04 waiting 5/6 ubuntu jujucharms 10 ubuntu
xenial-
...
[See attachment for full juju status output]
Changed in juju: | |
status: | Invalid → Triaged |
importance: | Undecided → Medium |
tags: | added: usability |
Also, when I removed the failed machines from the model, 'juju status' worked again with 2.1.2. The error content appears to be the source of the problem.