Closed samvantran closed 5 years ago
@mpereira yup, I think 30 minutes would've been enough too but opted for 60mins bc it was consistent with the infra step (it serves as a max timeout anyhow). From the log during the test, 15 mins was enough to deploy 46 of 50
20:39 (Sam Tran) the script timed out after 15minutes but 46 non-gpu dispatchers were deployed successfully
What changes were proposed in this pull request?
Resolves DCOS-58440
In MWT18, this script timed out after 15 minutes and we were not able to successfully deploy all dispatchers. This extends the timeout to 60 minutes so we give ample time for them to deploy. Note: 60minutes is the same time we give to deploy all the infrastructure (Kafka, C*, etc)
How were these changes tested?
This is part of MWTs