We are running Smartlist v3.15 and are running into a problem on Solaris 2.7. We have a mailing list with
around 15,000 names on it. Due to the downturn in the economy, we now have hundreds of bad addresses on the
list.
When a message goes out and we get hundreds of responses for bad addresses, Smartlist stops working. The
system is okay as long as there are only about 100 - 150 returned messages. After the system has more than
100 - 150 procmails running, multigram no longer runs and Smartlist stops processing the procmails.
This system has plenty of memory and swap space available. The number of procmails continues to grow
after Smartlist dies, and we have seen as many as 600 processes at one time. Once the system is rebooted,
Smartlist is okay again.
When Smartlist stops working we get these errors in the log after a while:
procmail: Terminating prematurely whilst waiting for lockfile "tmp.lock"
Does anyone have any ideas what might cause this? Is there a system parameter that needs adjusting?
Thanks,
Bill Endter