Join GitHub today
GitHub is home to over 20 million developers working together to host and review code, manage projects, and build software together.
handle reset timeouts better #45
Comments
Yrrsinn
commented
Jun 10, 2016
|
This uncaught KeyError exception is still around:
|
|
Bundletester resets the env between each test. If it can't complete the reset in 60s it gives up. I'm not sure why it's taking so long to remove the quobyte-registry service. I'm also not sure what the best way to handle it is. One idea is to allow the timeout value to be configured in tests.yaml, so you can increase it. Also need to handle the timeout more gracefully, but that's what this bug is already about. :) Bundletester tests charms and bundles. Part of the testing is detecting and executing Makefile targets. You can turn this off by passing |
|
Should've also mentioned: if you don't want to reset the env between each test you can add |
Yrrsinn
commented
Jun 10, 2016
•
|
Ah, thanks for pointing out this details of bundletester. Resetting between tests is fine, I will set-up a better integration for my makefile. Actually the tests for this charms does not always fails, sometimes all tests passes. I guess the error is less related to a time-out, but to some problem with Juju Storage. From the output of From the debug-log:
machine-3, 4 and 5 are the hosts quobyte-registry has been deployed to. machine-3 and 4 failed to die. machine-5 is gone. (Edit: additional note:
|
|
@Yrrsinn You may want to submit this info as a bug to juju-core: https://bugs.launchpad.net/juju-core |
tvansteenburgh commentedMay 10, 2016