All users were logged out of Bugzilla on October 13th, 2018
Steps to reproduce: 1. open https://treeherder.mozilla.org/#/jobs?repo=try&revision=8cdb2839cb8afa9563dee59236f2420678253bd7 2. click "down triangle" - "Add new jobs" 3. select SM(cgc) for windows xp opt 4. click "Trigger New Jobs" 5. click "OK" for "This will trigger all selected jobs. Click "OK" if you want to proceed." dialog Actual result: it transits to the same state as step 3 (at least visually), and it stays at the same state until the request is handled and green confirmation gets shown at the top left. usually the delay was several seconds, but today it's forever, and it's hard to figure out it's waiting or something goes wrong Expected result: it transits to some state that is different from step 3, and shows some progress that the request is now being handled, and then shows green confirmation gets shown at the top left that the request is handled.
Component: Treeherder → Treeherder: Job Triggering & Cancellation
I'm pretty sure we fixed a bug a while back about the error messages not being shown in certain cases. Have you seen this recently?
it's not about error messages. what I proposed here is to indicate some progress, after triggering new jobs, before showing "Request sent..." confirmation. today it took 17 seconds after clicking "OK" button in step 5, until the green "Request sent..." confirmation is shown. displaying "Sending request..." or "Waiting for response..." or something like that will clarify what's ongoing, and it tell me that I haven't accidentally cancelled triggering it, because otherwise the screen after step 5 is same as step 3, there "Trigger New Jobs" button is available. let me know if it's still unclear.
(In reply to Tooru Fujisawa [:arai] from comment #2) > it tell me that I haven't > accidentally cancelled triggering it, because otherwise the screen after > step 5 is same as step 3, there "Trigger New Jobs" button is available. maybe this was too-specific example. the same thing happens when some error happens, like, login expires or some network trouble happens. in any case the state after triggering should be different than before that, or after some issue happens and is waiting for re-trigger.
Ah I understand now - thank you for the explanation :-)
Perhaps we could either throw up a notification, or add a spinner to the actual button to indicate to the user that the action is in the pipeline.
Priority: -- → P2
You need to log in before you can comment on or make changes to this bug.