As we commented in the last call: 1) These are the empty tabs that can be deleted in Jenkins landing page: - Merge-Flourine (remove or repair) - Merge-Neon (remove or repair) - Merge-Nitrogen -

https://git.opendaylight.org/gerrit/#/c/80664/

CSIT jobs are failing due to python dependency not satisfied: https://jenkins.opendaylight.org/releng/view/bgpcep/job/bgpcep-csit-3node-bgpclustering-all-sodium/18/console 07:57:51tornado requires

Hi, Venkat I tried it, it did work, my question is why a real external network can’t work but simulated external network can work. By the way, snat can’t work with this simulated external

There does not seem to be anything out of the ordinary from CDS perspective, but there are some OLFE/validation errors. Can someone from the OFP team take a look, please? Thanks, Robert

FYI tell-based test with latest master: https://jenkins.opendaylight.org/releng/view/openflowplugin/job/openflowplugin-csit-3node-gate-clustering-only-sodium/9/ I still see topology link information

Ok so I’m proposing this: https://git.opendaylight.org/gerrit/#/c/80586/ Let me know your thoughts. Alexis

Hi Venkat, Let me ping Vexx to get a odl-standard-16 created. I had thought most of the flavors were available to us as odl-*. In the meantime you could use v2-standard-16 for testing. The difference

Hello Alexis, looks good. karaf4-parent is not the right place, as they are inherently CDS-aware. I think the best option would be to provide them through a separate artifact from

Greetings all, Following a call last week, I started creating a specific ODL distro for ONAP’s need. Here is the current draft: https://gerrit.onap.org/r/#/c/79075/ Question is: when creating the

Luis, It is a local change to integration-deploy-openstack-run-test.sh using pabot. I will submit a patch for review by thursday. -- Regards, Venkatrangan G

Thanh, It seems a very few flavors are names as "odl-". A majority of them are still v1- and v2-. I printed the flavors from os-cloud "vex". This is the list -[0]. I am trying the job with

I agree to Sam, that we should invest more time on stabilizing and making apex jobs as voting ones. Any other optimisations can come on top of that. Thanks, Faseela

Hi Sam/Jamo, Thanks for the detailed mail explaining the problems. Do you think it will be better to have just one job running tests in parallel and

If it (Thanh's idea of putting everything in containers) saves $$ but keeps the execution times similar - isn't that a benefit? Since that may reduce some infra costs?

Sam, I *think* Thanh's idea for putting everything in containers was more to the point of only having a single VM running for a CSIT job. so the compute, control and odl nodes run in containers and

+1 jamo, Overall, it seems pretty risky to muck around with running parallel suites. It might be better to optimize on what tests we want to run at high frequency vs daily vs weekly. Optimize any

Attention integration-devs, Autorelease neon-mvn35-openjdk8 failed to build odl-integration-all from integration in build 35. Attached is a snippet of the error message related to the failure that we

The main gain from containers is the reduced communication, if you ignore the spin where the real savings is. Meaning apex gets us the spin up savings with containers, but after that the main latency

my guess is that Venkat did some tweaks in his local releng/builder repo and created a sandbox job to try the parallel stuff, so maybe there is no upstream/gerrit code to look at. but, probably it's