What would be the benefit over the current approach? You can just fiter JIRA for issues with a specific label:
https://issues.apache.org/jira/browse/FLINK-2771?jql=project%20%3D%20FLINK%20AND%20resolution%20%3D%20Unresolved%20AND%20labels%20%3D%20test-stability%20ORDER%20BY%20priority%20DESC Not sure if the gain (if there is any) justifies the overhead of adding issues to an umbrella issue. 2015-10-09 13:47 GMT+02:00 Matthias J. Sax <mj...@apache.org>: > Hi, > > right now we collect instable tests via labeling them with > "test-stability". > > I think it would be a better approach, to use an Umbrella Issue in JIRA > and collect them as Sub-Tasks. Storm does the same thing: > https://issues.apache.org/jira/browse/STORM-915 > > This would have the advantage, that we get a simpler overview what tests > got already reported (and/or fixed). We can also list the tests by > component which makes it easier to find (the JIRA "search issues" > feature is cumbersome to use because the UI is quite slow) > > I would keep current approach and label the sub-tasks with > "test-stability" and mark them as "critical". > > What do you think about it? > > > -Matthias > >
