I don't think we're resource starved enough to have to resort to overcommit tricks. Let's just start a lot of instances and give everyone lots of room :) Lighty ones should be migrated later, but should move the current uwsgi ones to the generic nodes before they start getting a lot of usage.
Description
Description
Details
Details
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | None | T90534 Make toolforge reliable enough (tracking) | |||
| Declined | None | T91068 Set up a schedule for doing failover exercises for toollabs | |||
| Resolved | Andrew | T90542 Make sure that toollabs can function fully even with one virt* host fully down | |||
| Resolved | yuvipanda | T91065 Move uwsgi jobs to be run on generic hosts, retire uwsgi hosts |
Event Timeline
Comment Actions
Change 193558 had a related patch set uploaded (by Yuvipanda):
tools: Move uwsgi services also to generic nodes
Comment Actions
Change 193781 had a related patch set uploaded (by Yuvipanda):
tools: Add uwsgi support to generic webgrid nodes
Comment Actions
Change 193791 had a related patch set uploaded (by Yuvipanda):
toollabs: Remove uwsgi node definitions
Comment Actions
Deleted the puppet code, and also deleted the two instances. Need to remove the queue and the exec hosts from OGE.