-
Notifications
You must be signed in to change notification settings - Fork 2.6k
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.
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
Modified node-kubelet-flaky job and image spec #17687
Changes from all commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,15 @@ | ||
--- | ||
images: | ||
cos-stable1: | ||
image: cos-81-12871-119-0 # docker 19.03.6, current latest LTS. | ||
project: cos-cloud | ||
machine: n1-standard-16 | ||
metadata: "user-data<test/e2e_node/jenkins/gci-init.yaml,gci-update-strategy=update_disabled" | ||
tests: | ||
- 'Node Performance Testing' | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. (this is me not fully understanding how all of this works) Does this take precedence over the It feels like we're overloading/misdirecting by using the term "flaky" when really this job seems to be running perf tests There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. it works in combination with the focus on And it is overloading and misdirecting. But for some reason the e2e tests running in this current version of the job were never properly modified to run in that machine and have been running since the beginning in an n1-standard-1. My idea here was to explicitly specify within the image config that this whole fix is only for a handful of performance tests that were labeled as flaky becuase they were running in the wrong machine. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. It happens to be the only test marked flaky left in node_e2e. I think the idea of both There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. ok, I won't block on this one as long as there's followup to either rename this job to something like we currently have no way of verifying/linking whether all e2e tests are executed by at least one CI job, so I could understand if someone wants to preemptively keep a flaky job around even if there are no flaky tests There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
YES to all of this! probably split out a new job for this (or move the contents of the image config into the jobs running cenchmarks TBD with SIG Node). |
||
ubuntu: | ||
image: ubuntu-gke-1804-1-17-v20200605 # docker 19.03.2 / containerd 1.2.10 | ||
project: ubuntu-os-gke-cloud | ||
machine: n1-standard-16 | ||
tests: | ||
- 'Node Performance Testing' |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the existing job also runs for ubuntu, can we include that here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
that is a good point... my decision making for this was purely based on sig node conversations about which images to run tests in and the cos ones are a definite yes.
Everything else that isn't cos seems to be up in the air at this moment (?).
During a previous meeting there was also a mention of organizing things such that each job only ran 1 image.
Right now, kubelet-flaky is running tests on this ubuntu image as well
test-infra/jobs/e2e_node/image-config.yaml
Line 6 in dbd9fce
so asking for feedback here: should this version of kubelet-flaky run ubuntu as well? and if so, should it run the same ubuntu image that it has been using until now or another?
/cc @spiffxp @MHBauer @bsdnet @vpickard
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@alejandrox1 Let's also update the Ubuntu since those are two images running. After we got them green, per the meeting, we may add more.
I will start a topic in sig-node to discuss how to support multiple images in a scalable way.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't know who monitors the output of this (clearly nobody), nor do I know if it surfaces in a dashboard.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sweet, so im looking at all the ubuntu images for the ubuntu-os-gke-cloud project
and
ubuntu-gke-1804-1-17-v20200605
seems to be the latest one.As per monitoring this later on, we'll figure it out 😅