ram wrote:
I have a mailing list server where an application sends mails to list
members using /usr/sbin/sendmail ( on commandline)
When the mailing is going on I find that /usr/sbin/sendmail takes quiet
some time to return even for a trivial message
For eg.
sendmail -f [email protected] [email protected] < /etc/hosts
"quite some time" is how much? It should complete in a fraction
of secound...
I need to identify the bottle neck. Do I need better CPU , or better
harddisks. There seems to be plenty of RAM free though
..but if your system is very busy at the same time, it may require
some more time.
Is your system idle or overloaded at the time you perform the test?
How about commands such as free(1), vmstat(1), top(1) and so on?
If the system is NOT overloaded, there's nothing to slow down
/usr/sbin/sendmail to a noticeable level. It just writes the
message to a spool file and notifies postfix about it, nothing
more (it's not entirely accurate since several processes are
involved here, but it's not the point -- the point is that it
all boils down to just writing a queue file). On any modern
system it takes fraction of a secound.
But it is possible you've other problematic places in your
config. For example, referencing to an unreachable LDAP
server in /etc/nsswitch.conf, non-working DNS resolver and
so on.
/mjt
