We need to query Treeherder to get datapoints for talos jobs (via th-client probably) for the tests that were succcessful, failed, running etc. This will help in determining the jobs that we need to backfill, trigger in an automated fashion. A rough method signature: > perfherder.query(startRev, endRev, branch, platform, test) > # returns a {'rev': [val1, val2,...], 'rev+1': [...], 'endRev': [...]} > * startRev: earliest rev to use in the pushlog sequence > * endRev: last rev to use in the pushlog sequence > * branch: needed to get signature > * platform: needed to get signature > * test: needed to get signature

Bug 1192976 will likely make this much easier to implement.

the latest perfherder client can easily get a signature and data for the signature in a given time range. With little work we can get all this data.