Darko Brdareski

> @brdji Please make sure the CI accepts this PR first. I ran the `go mod tidy` command locally, and pushed the changes. Now I notice this error message in...

> https://circleci.com/gh/testground/testground/6563 is failing It's passing now. I have to point out that I have re-run the pipeline 8 times before it worked - we should open an issue to...

I have restored the code that applies the desired `sysctl` changes to run pods. Since we already modify `net.core.somaxconn` for the plan nodegroups when we deploy the cluster, these changes...

> I don't quite understand how we're actually testing wether the sysctls are set as we please or if the defaults are used. Could you reword or elaborate on it...

> > Isn't this - > ..... > - only executed if `sysctls` is set in `env.toml` file? Something like here: You are correct, I got some information mixed up....

I tried reproducing the fix (running `network/ping-pong` testcase), but the error message is still there: ``` May 18 14:25:40.223260 INFO run finished successfully {"run_id": "ca2g3o3pft4h10nihkf0", "plan": "network", "case": "ping-pong", "runner":...

I now understand that `INCOMPLETE` is unrelated to the actual job outcome, thanks! Yes, using `InitializedTestCase` does solve the issue, and the job's outcome is correctly shown as successful in...

That seems like a good solution. We should pay attention to document it properly, and the (breaking) change it introduces in the new sdk version.

As we have started wrapping up the EKS cluster deployment, I am closing this task. The cluster performance reports can be found here: https://github.com/testground/testground/issues/1410 More information on what the main...

All that is left is to fix whatever is causing issue #1416 . Once that is complete, we will prepare a PR for all the required changes and close this...