Random "Server Errors" which don't show up in logs, only on the dashboard statistics (and when you run into them)

[Per]

AppID is "small-improvements-hrd".

We are seeing plenty of server errors which leave no traces in the log files, but we and our users keep running into them, and it simply shows as "Server Error" written in black on a white empty screen.

This seems to have started about 6 to 8 hours ago. First we also had other server issues (causing 500 errors) but after restarting our servers since memory consumption seemed high, those 500-errors went away. Now we're stuck with server errors that we cannot properly identify and that are hard to report. Any ideas?






Like always, I'd like to point out that I wouldn't be posting such a trivial support request to the forums, but since I've rarely (never?) heard back from previous support requests, I guess this is the best way. See Brandon's post for stronger wording :)

Cheers,
Per

[Per]

I just read Christina's postmortem analysis on the Downtime forum.

It's interesting that the errors I reported above started around the 19th (when the configuration was changed) and ended on the 22nd (after the outage).

Who knows, maybe the entire outage could have been avoided by analysing our bug report in time :) Well, at least the error I described above is now gone, without any action on our end.

Cheers,
Per