stalwart mail pegs multiple CPU cores at 100% for extended periods of time...

[rcfa]

So here my little server has 24GB RAM, and four CPU cores. I’m in essence the only user of the system. Access via IMAP is from from four iOS devices and a single Mac, e-mail traffic is minimal (a few dozen mails per day). Total archived mail volume around 20GB (so it could essentially fit all in RAM).

Despite this rather light load, I catch the server rather often under high system load, sometimes multiple stalwart mail threads pegging all four corese near 100% CPU time. Acual RAM use is minimal, less than 25%, no swap.

What’s going on? Is this normal? How would I go about figuring out what’s happening?
There’s of course the occasional attempt to spam or hack, but I can’t find anything in the log files substantial enough to explain such loads.

[mdecimus]

This is not normal. To diagnose I'll need a gdb trace. Next time you see high CPU usage execute as root:

$ gdb -p <PID>
set pagination off
set print thread-events off
thread apply all bt

Where <PID> is the Stalwart process id. Once you have the output please paste it here.