Rietveld Code Review Tool
Help | Bug tracker | Discussion group | Source code | Sign in
(636)

Issue 55900043: Refactored Juju Run

Can't Edit
Can't Publish+Mail
Start Review
Created:
11 years, 11 months ago by waigani
Modified:
11 years, 11 months ago
Reviewers:
mp+202792, thumper
Visibility:
Public.

Description

Refactored Juju Run Proxy env vars are passed into unit context i.e. `[ -f "/home/ubuntu/.juju-proxy" ] && . "/home/ubuntu/.juju-proxy"` is prepended to juju run cmd inside RunCommand.executeInUnitContext method https://code.launchpad.net/~waigani/juju-core/refactor-juju-run/+merge/202792 (do not edit description out of merge proposal)

Patch Set 1 #

Unified diffs Side-by-side diffs Delta from patch set Stats (+7 lines, -4 lines) Patch
[revision details] View 1 chunk +2 lines, -0 lines 0 comments Download
cmd/jujud/run.go View 1 chunk +3 lines, -1 line 0 comments Download
state/apiserver/client/run.go View 1 chunk +1 line, -2 lines 0 comments Download
state/apiserver/client/run_test.go View 1 chunk +1 line, -1 line 0 comments Download

Messages

Total messages: 4
waigani
Please take a look.
11 years, 11 months ago (2014-01-23 06:20:01 UTC) #1
thumper
These look good. We were also going to check the err response from the fslock ...
11 years, 11 months ago (2014-01-23 20:55:19 UTC) #2
waigani
On 2014/01/23 20:55:19, thumper wrote: > These look good. > > We were also going ...
11 years, 11 months ago (2014-01-27 09:27:38 UTC) #3
thumper
11 years, 11 months ago (2014-02-12 02:24:05 UTC) #4
On 2014/01/27 09:27:38, waigani wrote:
> On 2014/01/23 20:55:19, thumper wrote:
> > These look good.
> > 
> > We were also going to check the err response from the fslock Lock method.
> 
> In run_test.go I've written:
> 
> func (s *RunTestSuite) TestGetLock(c *gc.C) {
> 	_, err := getLock()
> 	c.Assert(err, gc.ErrorMatches, "mkdir /var/lib/juju/locks: permission
denied")
> }
> 
> Is this what you meant? It passes, but I don't know if  "mkdir
> /var/lib/juju/locks: permission denied" is expected. I don't understand the
> problem / remember why/how we were going to test the fslock?

Yes that is expected, and there is already another bug outstanding for that.

LGTM
Sign in to reply to this message.

Powered by Google App Engine
RSS Feeds Recent Issues | This issue
This is Rietveld f62528b