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
[Networking] [E2E] n-to-1 iperf client->server throughput benchmarking #22869
Conversation
@kubernetes/sig-testing a first pass ateempt at what we discussed tuesday, using iperf, which measures netowrk performance.
|
30da5fe
to
e7d5fa7
Compare
We found a Contributor License Agreement for you (the sender of this pull request) and all commit authors, but as best as we can tell these commits were authored by someone else. If that's the case, please add them to this pull request and have them confirm that they're okay with these commits being contributed to Google. If we're mistaken and you did author these commits, just reply here to confirm. |
CLAs look good, thanks! |
I've now run this on GCE several times and its stable. A Test result for 3 nodes follows -- this is done on GCE. I've also tested this on OpenShift a quick first test it seemed to work similarly (and give reasonably numbers). [begin] Node,Bandwith CSV |
d763e17
to
11319d1
Compare
Removed the do-not-merge label, and ran a few tests of this on GCE. It stable from what I can tell, but I labelled as flakey since it is a new test which we will want to bake into the CI over time. |
9b545f9
to
89a607f
Compare
@@ -527,6 +601,8 @@ ReturnPodsSoFar: | |||
// Next: Pod must match at least one of the states that the user specified | |||
for _, pod := range unfilteredPods { | |||
if !(passesPhasesFilter(pod, p.ValidPhases) && passesPodNameFilter(pod, p.PodName)) { | |||
// This can be used for debugging but otherwise is very verbose. | |||
// Logf("phase or name not matched [[[ %v ]]] : %v %v ", pod, p.ValidPhases, p.PodName) |
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.
just delete it? or add a slimmer version
This looks fine for a first cut as long as we move it to feature so it doesn't block anything, and the nits are addressed |
@bprashanth comments addressed
re: iperf vs netperf vs ... they're all functionally equivalent for the use cases we care about i think. thanks for the review ! |
generic pod-per-node functionality for testing - 2 node test only - update framework to decompose pod vs svc creation for composition. - remove hard coded 2 and pointer to --scale
GCE e2e build/test passed for commit 95e315e. |
Hi @bprashanth any other comments? |
In that case we already have a netperf tester? LGTM |
Removing LGTM because the release note process has not been followed. |
@bprashant can we fix this label , thanks.! |
And yeah, netperf as a container can be swapped in here. But netperf-tester you linked appears to be measuring network performance for a simple one off scenario . This test attempts to scale with respect to cluster size and create replication controllers both for servers and clients... |
@k8s-bot test this Tests are more than 48 hours old. Re-running tests. |
GCE e2e build/test passed for commit 95e315e. |
GCE e2e build/test passed for commit 95e315e. |
@k8s-bot test this [submit-queue is verifying that this PR is safe to merge] |
GCE e2e build/test passed for commit 95e315e. |
Automatic merge from submit-queue |
Automatic merge from submit-queue Logging soak Implements #24427 Needs - #24471 so that it doesnt clog test outputs for scale - builds on the utils function added in support of #22869 cc @timothysc @kubernetes/sig-testing
No description provided.