Hello,
Le jeu. 24 sept. 2020 à 04:21, Larry Shen larry.shen@nxp.com a écrit :
Hi, Remi,
We found only 2 sites have this issue, other sites ok.
After double check, they use something like “lavacli -i admin@validation jobs wait --polling 60 $job_id”
As temporary solution, already ask them to reduce the RPC call with “—polling 300”
Look like the server is struggling to cope with the load.
Looks no this issue for last 1 week already, so I will pending this check.
BTW, I remember we could monitor the event of job finish, will this remove the polling of master issue?
Will it also work after this https://git.lavasoftware.org/lava/lava/-/merge_requests/1253 ?
Yes the zmq event stream is still available. It's also possible to use web sockets to get the event stream.
Rgds