Hi all, Over the last week, we finally managed to get a better handle on some of the CPU spin-locks some people have reported.
I thought I'd share the simple shell script that helped me generate the kind of information I needed. http://git.dbmail.eu/paul/dbmail/tree/test-scripts/dumpbt.sh If you take it, make it executable, and run it when your dbmail get's stuck a a high cpu load, it will produce simple backtraces for the threads involved. Since I can't be certain the fix earlier this week covered all such scenarios, it might be useful for others. doing: top -H -p `pidof dbmail-imapd` will give you a quick overview of what's going on. If it's the main thread, the second process with command 'dbmail-imapd' will be at a high cpu load. If it's one or more worker thread, the command will be 'pool'. you can attach a gdb to the process involved and do a backtrace. Interactively that would entail: gdb -p $PID (gdb) bt where $PID is the process or thread-id you wish to investigate. The script I'm referring to above simply automates this. -- ________________________________________________________________ Paul J Stevens pjstevns @ gmail, twitter, github, linkedin www.nfg.nl/i...@nfg.nl/+31.85.877.99.97 _______________________________________________ DBmail mailing list DBmail@dbmail.org http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail
