Join GitHub today
GitHub is home to over 20 million developers working together to host and review code, manage projects, and build software together.
worker/uniter/runner/debug: better test logging #6402
Conversation
| @@ -0,0 +1,11 @@ | ||
| +// Copyright 2013 Canonical Ltd. |
| @@ -40,6 +44,9 @@ func (s *ServerSession) RunHook(hookName, charmDir string, env []string) error { | ||
| cmd.Env = env | ||
| cmd.Dir = charmDir | ||
| cmd.Stdin = bytes.NewBufferString(debugHooksServerScript) | ||
| + var combinedOutput bytes.Buffer |
axw
Oct 7, 2016
•
Edited 1 time
-
axw
Oct 7, 2016
Member
Edited 1 time
-
Oct 7, 2016
This means a hook's output won't show up until it completes, which can take an arbitrarily long amount of time. This may have an impact on usability of debug-hooks.
Please either add Stdout and Stderr parameters to HookContext/ServerSession and use them, or modify echocommand to do what you need.
howbazaar
Oct 7, 2016
Owner
I've reworked how we get the output for the test now, can you take another look plz?
|
$$merge$$ |
|
Status: merge request accepted. Url: http://juju-ci.vapour.ws:8080/job/github-merge-juju |
jujubot
merged commit ffd6121
into
juju:master
Oct 7, 2016
howbazaar
deleted the
howbazaar:better-test-logging
branch
Oct 7, 2016
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
howbazaar commentedOct 7, 2016
This work was in response to a test timeout in the debug package.
After much examination it was determined that the test timed out because RunHook failed to start properly, and the case statement in the first select was only calling c.Error. Error only marks the test as failing, it doesn't actually cause the test to stop, so when it did fail there, the test kept looping.
Extra testing.LongWait calls added to each select as a firm backdoor. Now the ServerSession captures the output of the hook execution in order to allow the test to write it out in the failure case.