Created attachment 251337 [details] query debug info The Summary is a dummy, we can fix it to be more accurate once we figure out what the issue is. Been having a lot of people mail in "Lost connection to MySQL server" errors since the upgrade. Sometimes they seem random. This one I can reproduce at will, so I'm assuming the query is taking longer to run than the timeouts are set for. Debug info attached.
How long does the query take if you run it actually in the MySQL console?
I forgot, here's a URL to the query: https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&classification=Client+Software&product=Mozilla+Application+Suite&product=Thunderbird&long_desc_type=substring&long_desc=en-us&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&keywords_type=allwords&keywords=&bug_status=RESOLVED&bug_status=VERIFIED&bug_status=CLOSED&resolution=FIXED&resolution=INVALID&resolution=WONTFIX&resolution=DUPLICATE&resolution=WORKSFORME&resolution=EXPIRED&resolution=MOVED&emailassigned_to1=1&emailtype1=exact&email1=&emailassigned_to2=1&emailreporter2=1&emailqa_contact2=1&emailtype2=exact&email2=&bugidtype=include&bug_id=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Bug+Number&field0-0-0=noop&type0-0-0=noop&value0-0-0= And here's how long it takes to run on the server: 36672 rows in set (11.26 sec) It takes a long time (2 or 3 minutes) to send that amount of data back to me in the Terminal window though. Perhaps it's hitting max-allowed-packet and giving a strange error message? Or if it is all coming in one packet maybe DBI is giving up waiting for it to transmit? Interestingly enough, that URL doesn't time out for me now, so maybe this is something transient...
OK, I was told recently that MySQL has a total lifetime timeout on connections, meaning any given connection cannot last longer than the timeout, even if it's not idle. Supposedly this timeout defaults to 8 hours. I can't seem to find it documented anywhere, however. But given that we're using mod_perl, and connections are being pooled, it's quite possible we've been hitting this timeout as apache processes grow to be that old. Perhaps we need to set a time limit on how old an apache process can get, or keep track of how long our MySQL connections have been open and kill them off ourselves before they get that old.
i like my testcase better, as the query that's failing is trivial, and the problem is that a previous query combined with expensive charting has caused the simple query to be unfortunately aborted.
Isn't this fixed by bug 408766?
Yes, this bug and that one sound identical, to me.
Status: NEW → RESOLVED
Last Resolved: 10 years ago
Resolution: --- → DUPLICATE
Duplicate of bug: 408766
You need to log in before you can comment on or make changes to this bug.