Closed Bug 1185276 Opened 11 years ago Closed 11 years ago

Fetching crash reports takes too long

Categories

(Socorro :: General, task)

Product:
Socorro
Component:
General
Type:
task
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED DUPLICATE of bug 1178475

People

(Reporter: kats, Unassigned)

Details

IF I have a a crash id, e.g. bp-1378954d-2ce8-41cc-8cd6-32e132150718 and I go to load it, I get a screen that says "Fetching this archived report will take 30 seconds to 5 minutes" which refreshes every 30 seconds. Often I have to wait many minutes for the report to show up and sometimes it doesn't ever appear. This is a pretty bad user experience and significantly impacts diagnosing and fixing crashes. Specially when the crash report is a recently submitted one it shouldn't be an "archived report", whatever that means.
I agree that this UI is terrible! I filed bug 1178475 on it recently. We also just had a processor outage this morning (bug will be filed momentarily), bp-1378954d-2ce8-41cc-8cd6-32e132150718 is there now. I am going to dupe your bug against the UI improvement bug - normally we process crashes within about 3-5s so making people wait at least 30s with a really confusingly-worded loading screen is not wonderful. Just FYI when you see this loading screen it means that your crash has not been processed (Socorro only processes 10% of release crashes so it could be that, or it could be that you're too quick ;) ), and is going into the "priority processing" queue so it's hopping the "normal processing" line to the front.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → DUPLICATE
This morning's processor outage is bug 1185278.
Thanks for the quick response and duping. If it is indeed a case of "crash has not been processed" (which is likely as they were very recent crashes) detecting that scenario and updating the wording would go a long way towards improving the UX.
The alerting has been fixed. In a somewhat inexplicable move, datadog gets no data and doesn't consider that to be < 1 processed. I set the flag for 'alert on missing data', and we should now receive that notification.
You need to log in before you can comment on or make changes to this bug.