Using this as my tracking bug.
<rdar://problem/38033914>
This was duplicated: https://bugs.webkit.org/show_bug.cgi?id=183218 https://bugs.webkit.org/show_bug.cgi?id=183249
This bug wasn't duplicated. Looking at logs of the other two... it appears we called migrate-bugs twice in a row. I'm not sure why.
for 183218, here is what I think happened: Looking at history: wilander@apple.com 2018-02-28 11:09:25 PST CC webkit-bug-importer@group.apple.com wilander@apple.com 2018-02-28 11:09:28 PST Assignee webkit-unassigned@lists.webkit.org wilander@apple.com webkit-bug-importer@group.apple.com 2018-02-28 11:09:54 PST Keywords InRadar see the importer was CC'd at 11:09:25 , and then another change was made to the bug (at 11:09:28) before the "InRadar" keyword made it back into Bugzilla (11:09:54).
the same thing happened to 183249: wilander@apple.com 2018-03-01 11:04:10 PST CC webkit-bug-importer@group.apple.com wilander@apple.com 2018-03-01 11:04:19 PST Assignee webkit-unassigned@lists.webkit.org wilander@apple.com webkit-bug-importer@group.apple.com 2018-03-01 11:04:38 PST Keywords InRadar
It looks like there is a race condition. If a bug is modified after CC'ing the importer, but before we get the InRadar keyword, we span another update attempt.
I'm not sure the best way to address this. We could consider a 'settle timer', to give bugzilla a chance to catch up before processing again.
I had a feeling it was some kind of race condition. I typically do it that way — CC the importer, then assign the bug to myself.