Skip to content
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

DBS configuration for NATS integration #821

Closed
wants to merge 1 commit into from

Conversation

vkuznet
Copy link
Contributor

@vkuznet vkuznet commented Dec 3, 2019

No description provided.

@muhammadimranfarooqi
Copy link
Contributor

@vkuznet , Valentin

Should I merge this PR?

Best
M Imran

@vkuznet
Copy link
Contributor Author

vkuznet commented Dec 10, 2019

Not yet, it is related to dmwm/DBS#617, I'm still awaiting for ok from Yuyi (@yuyiguo )

@yuyiguo
Copy link
Member

yuyiguo commented Dec 10, 2019

I haven't got time to review the DBS PR, but why this PdmV related to DBS? I won't have time to review and test the DBS one for a while. @vkuznet, can you separate them?

@vkuznet
Copy link
Contributor Author

vkuznet commented Dec 10, 2019

Yuyi, I'm not sure I understood your comment. The dmwm/DBS#617 provides integration with NATS, this PR only provide how to configure DBS to use or not NATS. In this PR I explicitly disable NATS such that default DBS configuration will not use it and it serves as an example how configuration should be done.

Regarding PdmV use case, it is listed here: https://its.cern.ch/jira/browse/CMSMONIT-161, they want to monitor datasets info (e.g. change of dataset status and growths of datasets).

@yuyiguo
Copy link
Member

yuyiguo commented Jan 23, 2020

@vkuznet
Following this configuration and the server info you gave me, I did not receive NATS message. I don't know why and have no time to debug this for this release. I have to move on to other things as we discussed on the phone.
Thanks,
Yuyi

@vkuznet
Copy link
Contributor Author

vkuznet commented Jan 23, 2020

Yuyi, I don't know when you performed tests, but today CERN network struggles completely, see full message https://hypernews.cern.ch/HyperNews/CMS/get/cernCompAnnounce/1393/1.html We had issues with frontends, CRAB, Monitoring, Cephfs, etc. And I don't know if CERN IT fixed the problem yet.

Said that I checked nats server logs and I saw records from dbs3-test2. Therefore DBS was talking to NATS server and I checked DBS logs on dbs3-test2 and there is no NATS errors either. Therefore it is probably a subscriber. I setup one for DBS in monitoring cluster now, therefore if your test instance is running and you'll do some dataset manipulation (e.g. change dataset status or inject new blocks/datasets), then I'll check the subscriber and report.

If you'll do any tests (apart from NATS) with dataset changes just report me a timestamp that I can check the logs.

@yuyiguo
Copy link
Member

yuyiguo commented Jan 23, 2020

Thu Jan 23 19:10:43 CET 2020
I did a write unit test that has a lot of writing.

@vkuznet
Copy link
Contributor Author

vkuznet commented Jan 23, 2020

Yuyi, I think we have invalid DBS configuration (may be my mistake, I think I posted in PR the original config I was trying to use).

The config you have defines NATS parameters in view section (the last lines of config), while it should be part of dbs section. In my k8s configuration I have this

# NATS integration
config.dbs.nats_server=xxxxx
config.dbs.use_nats=True
config.dbs.nats_topics=[]

instead of parts in view section of your configuration. So the NATS configs goes to config.dbs part instead of dbconf as you have now.

Can you make this change and re-run unit tests?

@vkuznet
Copy link
Contributor Author

vkuznet commented Jan 23, 2020

Yuyi, also according to DBS logs on dbs3-test2 the NATS manager was never called. According to the code https://github.com/dmwm/DBS/blob/master/Server/Python/src/dbs/web/DBSWriterModel.py#L43-L51 we should have an entry in DBS logs. But it never happens. That's why I think we improperly configured DBS.

Since I never understood to full details DMWM configuration system I always struggle to find out proper place in it for additional stuff.

@vkuznet
Copy link
Contributor Author

vkuznet commented Jan 23, 2020

I can modify this PR if necessary with the change I mentioned in previous comments, but it would be nice if you'll prove in your environment/VM the proper configuration place for NATS attributes.

@yuyiguo
Copy link
Member

yuyiguo commented Jan 24, 2020

Valentin,

I don't have time to try again today. The preprod are having problem and I don't think we have time for this release.

Thanks,
Yuyi

@vkuznet vkuznet closed this Feb 21, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants