This may help with test level SETA, or maybe just arranging the chunks so failures are concentrated. First, is a simple list of every test and the last time it was seen failing . Combined with the first time the test was seen , we have a small hint of the stability of the test (and how far back ActiveData currently goes)  http://activedata.allizom.org/tools/query.html#query_id=OZ_9uTBY  http://activedata.allizom.org/tools/query.html#query_id=YYq4fJVw
this is neat, I pulled up the query and it just had stuff that failed in the last week or so, not much data. Maybe we could reverse sort it since it only shows 20 or so items?
Yes, it definitely needs more data; hopefully I have the daemon (bug 1193250) filling the cluster with the older data. Also, these two queries need to be tied with higher-level logic/program: The first query only pulls the ones that failed, but the second is the full list of tests.
You need to log in before you can comment on or make changes to this bug.