🚀 go-pugleaf

This is bizarre - perhaps someone's seen it before:

Every once in a while (usually right after a reboot), I get sendmails
syslogging "Alias database out of date", and then (apparently) going
into an infinite loop (at least, the processes are picking up a *lot*
of CPU time).  Queue runs never exit, and eventually the machine grinds
to a virtual halt due to the load of all the looping sendmails.  gcore-ing
the offending processes and doing a stack traceback reveals that the
call stack does *not* lead back to main(), which leads me to suspect
that the stack is being corrupted.  Running "newaliases" does not help;
the only way out is to kill off all the sendmails, run "newaliases",
and restart the sendmail daemon.

If it makes any difference, we run a "sendmail -bz" and "sendmail -bi"
in rc.local before starting the sendmail daemons (our machines tend
to change their hostnames a lot, and the hostname is wired into the
frozen config file - refreezing in rc.local is meant as a workaround
for this).

Anyone seen anything like this before?

Thanks,
--Pat.
-- 
Pat Barron
Internet:  pat@orac.pgh.pa.us  - or -   orac!pat@gateway.sei.cmu.edu
UUCP:  ...!uunet!apexepa!sei!orac!pat  - or -  ...!pitt!darth!orac!pat