I have not received a answer yet, but I am posting this summary to tie up
the loose ends before I go on vacation Sunday (Yea!!!).
I have had questions replies from ...
deb@ATWC.Teradyne.COM
rlk@Think.COM
We thought it might be a kernel bug, but the process is easy to kill. I
have a core file, but dbx says ...
ack# dbx /usr/lib/sendmail.mx /var/spool/mqueue/core
warning: cannot read pcb in core file: registers' values may be wrong
warning: no symbols
and waits here forever. I am not familiar with adb, but I tried it and got ...
adb /usr/lib/sendmail.mx /var/spool/mqueue/core
core file = /var/spool/mqueue/core -- program ``sendmail.mx''
core file ("/var/spool/mqueue/core") appears to be truncated.
SIGBUS 10: bus error
data address not found
any ideas how to proceed??
> From don Wed May 5 09:07:51 1993
>
> Long lines, needed stuff at RHS of ps output.
>
> Our network has been VERY stable lately, and a problem has arised.
>
> It seems that when a client dataless workstation has been up for a
> long time, that sendmail hogs memory.
>
> For instance, for a host here, ps auxww reports ...
>
> USER PID %CPU %MEM SZ RSS TT STAT START TIME COMMAND
> root 15480 0.0 5.344088 1608 ? S Mar 1 0:16 -(null) To mars (sendmail.mx)
>
> ps alxww reports ...
>
> F UID PID PPID CP PRI NI SZ RSS WCHAN STAT TT TIME COMMAND
> 20488201 0 15480 15479 0 15 044088 1604 kernelmap S ? 0:16 -(null) To mars (sendmail.mx)
>
> uptime reports ...
>
> 8:59am up 107 days, 19:14, 3 users, load average: 1.05, 0.90, 0.54
>
> now why on earth is sendmail using so much memory?
>
> PS, we found this error when clients started to run slowly, and then had
> memory shortages.
>
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:49 CDT