Yes, for one week the time for executing all CSIT in releng may increase ~20% because of the resource reduction.

Does this mean all jobs needing a robot minion will be impacted?

You have my vote for 1) and 20% reduction. Luis, you said you have a netconf perf job that we can use to validate this right? Some job that never fails when run during idle times or locally, but seems

Hi everybody, Yesterday we discussed in the integration call the problem of unstable perf/scale tests due to a possible infra congestion, e.g. all CSIT tests trigger at once after an AR distribution

just fyi, the new cli to use is 'robot'. Looks like a RF release that happened just a few hours back brought this change. patch to update our CSIT is

Hi Everyone, Not feeling well today so will be missing the meeting today. A few updates from me: Luis merged my inttest patches https://git.opendaylight.org/gerrit/#/q/topic:inttest-macros please

Ah, okay, I didn't realize it's expected for Jenkins to vote in the code-review field for unstable jobs. That put me in a prospective of also doubting the +2, since I expected a non-author's review to

Ah, okay, I didn't realize it's expected for Jenkins to vote in the code-review field for unstable jobs. That put me in a prospective of also doubting the +2, since I expected a non-author's review to

Hi Daniel, I don't see anything wrong here. Your +2 vote is satisfying the Non-Author check so things are normal here. As for Jenkins voting -1 in CR. This happens when a job is UNSTABLE rather than

Hello, I'm seeing some very strange Gerrit voting bugs in the last few hours. * My +2 on a patch I have nothing to do with is in Code-Review, not Non-Author-Code-Review, flag * Jenkins verify -1 is

Hi Sam, That's correct. The non-custom versions of the csit jobs do not let you set testplan because we removed it from defaults.yaml and reverted the default back to the job-template. The "custom"

That's right, we're applying some best practices we learned while building global-jjb. The intent here is as we continue these optimizations it will make the CSIT jobs much easier to read and

After looking at the scale tests in detail (item 3 below) I think they are failing because of infra (not controller) issues. FYI I run the same scale tests few times in my local env as well and in the

Yes, it is currently set in the project yaml. But I think with the recent restructuring of the templates that the variable "testplan" is not longer used in teh basic inttest-csit-1node - that template

Hi Sam: Cheers, Anil

Greetings Prateek: I've the code in the github repository here [1.] for drawing the ODL project charting and [2.] has the sample charts that I generated for ODL Carbon release. You'r converting the

I have created the dependencies graph for the flouring dependencies in an automated fashion using python graphviz. Just we need to feed in the dependency log file and we can get the output as a

To everybody, I went through Thanh’s patches to optimize CSIT jobs definitions and merged all of them so he can continue with the

Thanh, to set a testplan via a job yaml, do we need to use the custom templates? Take the coe-csit job [1], the yaml sets testplan:coe-container-networking.txt, but it is using the inttest-csit-1node

I've only looked at a handful of jobs that is spawned by patch-test and it doesn't appear like GERRIT_REFSPEC is used by them so might be safe to use for your uses. Might want to double check for the