Thanks Mark.
In the case of the server's own name, it's probably best to make sure that /etc/hosts is set up properly. Note that the FQDN *must* be first; many Solaris systems are set up with a non-FQDN first.
Yes, the server's own name is fine and everything in /etc/hosts looks to be correct. It seems to just be a problem with the reverse lookup of the Sprint client IP address.
In your case, if you aren't in a position to fix the client's DNS address, you may be forced to turn off reverse DNS lookups. I advise against doing so except as a LAST RESORT; it will make the logs less useful and it will break Kerberos authentication.
Ok - thanks for the advice. Unfortunately, I do not have any direct control over the DNS servers we use. Those are controlled by MIT's main IS&T folks. I am able to do a lookup from home so the problem doesn't appear to be the clients/Sprint's but the route that MIT is taking to do the lookup so I guess I should first contact them regarding this and see if something is misconfigured on their servers.
In the meantime, I'll try compiling it on a test server with allowreversedns set to NIL and see if the problem is fixed. It also seems to be that running a local caching name server may also aid in speeding up the server and the lookups. Is this something recommended? We currently do not do this on our Solaris servers but it was on the to do list. It seems the IMAP/POP server's performance might improve a little if this were done.
thanks much, --tom _______________________________________________ Imap-uw mailing list [email protected] https://mailman1.u.washington.edu/mailman/listinfo/imap-uw
