Create a pool of new workers for performing recurring releng tasks

RESOLVED FIXED

Status

Taskcluster
Worker
RESOLVED FIXED
2 years ago
2 years ago

People

(Reporter: coop, Unassigned)

Tracking

(Blocks: 1 bug)

Details

(Whiteboard: [docker-worker])

(Reporter)

Description

2 years ago
We need a new worker type on which to run recurring releng tasks. This is meant as a replacement for cronjobs running on cruncher (bug 1235851).

There should be a small pool of these workers to provide some concurrency for tasks. We should also have workers in all the AWS regions we currently run in for resiliency.
We create and destroy instances on demand, so there's no need for a pool. And the workers are dynamically created in the cheapest region, so we are resilient to region failure as well.

I created a workerType named "releng-task" with a maximum capacity of 20 (simultaneous tasks) on m4.large's with a max of two tasks per instance.  I also granted queue:{create,define}-task:aws-provisioner-v1/releng-task to the `releng` LDAP group, so you should be able to do so.
Whiteboard: [docker-worker]
Component: Docker-Worker → Worker
Status: NEW → RESOLVED
Last Resolved: 2 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.