Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Regression in e2e tests. #8499

Closed
timothysc opened this issue May 19, 2015 · 8 comments
Closed

Regression in e2e tests. #8499

timothysc opened this issue May 19, 2015 · 8 comments
Labels
area/test-infra priority/important-soon Must be staffed and worked on either currently, or very soon, ideally in time for the next release.

Comments

@timothysc
Copy link
Member

As of last Friday our continuous integration started failing. We purposefully run our continuous integration in a different configuration then upstream does in order to sus-out differences in deployment etc.

1 definite issue is --host is now failing through the e2es possibly due to changes in kubeclient utils.

/cc @rrati @roberthbailey @jayunit100 @quinton-hoole
@timothysc
Copy link
Member Author

rollup of new failures that all previously passed, when running on host = master
http://fpaste.org/223340/43204412/ sans the DNS tests.

rrati pushed a commit to rrati/kubernetes that referenced this issue May 19, 2015
The e2e tests will re-honor --host option. kubernetes#8499
d9c572d
@timothysc
Copy link
Member Author

@pmorie on inspection of logs on a recently updated Atomic 7.1.2 cluster w/Docker 1.6.0 we're seeing the following failure on the emptyDir tests:

"Unable to mount volumes for pod "pod-cf5888d8-fe31-11e4-b097-ecf4bbc72674_default": invalid argument"

@pmorie
Copy link
Member

pmorie commented May 19, 2015

After some debugging on IRC, it looks like the mount is failing because the rootcontext mount option is malformed:

pod cf58a9a7-fe31-11e4-bdb0-ecf4bbc72674: mounting tmpfs for volume test-volume with opts rootcontext="system_u:object_r:var_lib_t:s"

The context comes directly from libcontainer.Getfilecon -- it looks like this is a bug with that function. I will open a libcontainer issue for this.

@ghost ghost mentioned this issue May 19, 2015
Merged
rrati pushed a commit to rrati/kubernetes that referenced this issue May 19, 2015
The e2e tests will re-honor --host option. kubernetes#8499
b7486b0
@roberthbailey roberthbailey added the priority/important-soon Must be staffed and worked on either currently, or very soon, ideally in time for the next release. label May 19, 2015
@timothysc
Copy link
Member Author

/cc @eparis

dchen1107 added a commit that referenced this issue May 21, 2015
@dchen1107
Merge pull request #8500 from rrati/e2e-host-option-fix-8499
df6c670 
The e2e tests will re-honor --host option. #8499
@ghost ghost mentioned this issue May 21, 2015
Closed
@jayunit100
Copy link
Member

Also we need to roll back the minPods=1 option that got added... Some clusters don't necessarily run pods before starting e2es

@timothysc
Copy link
Member Author

Related Ticket:
#8534

Issues we've found:
#8666
relative path break on: bf1de72
I believe there are a couple of other relative path issues on update-demo from other commits.

It looks like there is implied semantic change that the tests expect to be run from
kubernetes/test/e2e

@timothysc
Copy link
Member Author

ok with #8666, we are green again with some caveats and possible workarounds for path issues. We have other tickets open around the SELinux issue so closing this one.

@bgrant0607
Copy link
Member

cc @dchen1107

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
area/test-infra priority/important-soon Must be staffed and worked on either currently, or very soon, ideally in time for the next release.
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
6 participants
@timothysc @pmorie @jayunit100 @bgrant0607 @roberthbailey @yujuhong