Scott R. Burns writes:
 > Some more interesting information.
 > 
 > 1) I have the pppd link up, routed -q running.
 > 
 > 2) I have removed resolv.conf. 
 > 
 > 3) Now I can ping the lan hosts very quickly (by ip addr). No pause before the ping starts. I think it must have been trying to reverse lookup the ip address specified to ping.
So the long-response problem is with DNS - do you have your local domain
primary zone files (forward and reverse) setup correctly, with
forwarders line in named.boot file? 

 > 
 > 4) I cannot check the response on the lan side as I am not there and all of those hosts are PeeCee's so I can't
 > telnet into them to check ;-(

 > 5) I think the popper was trying to lookup the ip addresses if the machines that were pop3'ing in from the lan and that is why that is slow. I guess with pppd down the gethostbyname() call would fail more quickly and thats why performance would be fine with pppd down ?
 > 

Yes - DNS lookups are probably falling more rapidly back to /etc/hosts
(assuming that you have maps there)

 > 6) The problem is I need resolv.conf up as I have squid running on the machine, and when the broswer requests come in from the clients squid requires dns to perform the name lookup. Because of this will I need to have DNS setup on the machine for the local lan and have any requests not resolved there forwarded to my providers DNS ?
 > 

Yes this is required (see above).

 > 7) Even with this configuration (resolv.conf renamed to hide it) telneting in from the outside world gives me a login prompt but after entering a username a password prompt never appears ? But rsh'ing in works fine ?

I think (not verified, rusty memory) that rsh does not do hostname
lookups; login does.

 > 8) It would appear that the default route entry is not working as if i ping a host out on internet the dns lookup works out to the provider dns servers but traceroute says it is going to send the data to the lan interface ?
 > 

The last entry in your routing table is suspect (assuming that
204.191.68.2/204.191.69.201 are local/remote addresses of PPP link). I
would have expected pppd to create

  default            204.191.69.201     UGS         4     3358      -  ppp0
  204.191.69.201     204.191.68.2       UH          0        0      -  ppp0

instead of:

  default            204.191.69.201     UGS         4     3358      -  ppp0
  204.191.68.2       204.191.69.201     UH          0        0      -  ppp0
  204.191.69.201     127.0.0.1          UH          1        0      -  lo0
        
Does your pppd options file have the local/remote IP addresses reversed?

 > www# traceroute www.netcontech.com (which is: 209.89.182.66)
 > traceroute: Warning: Multiple interfaces found; using 192.168.1.4 @ ep0
 > traceroute to www.netcontech.com (209.89.182.66), 30 hops max, 40 byte packets
 > 
 > It should have used the ppp0 interface as it is marked as the default route ?
 > 
 > www# netstat -r -n
 > Routing tables
 > 
 > Internet:
 > Destination        Gateway            Flags     Refs     Use    Mtu  Interface
 > default            204.191.69.201     UGS         4     3358      -  ppp0
 > 127.0.0.1          127.0.0.1          UH         16      107      -  lo0
 > 192.168.1          link#1             UC          0        0      -  ep0
 > 192.168.1.4        00:a0:24:85:a3:56  UHL         1       32      -  lo0
 > 192.168.1.5        00:a0:c9:90:23:fd  UHL         0       33      -  ep0
 > 204.191.68.2       204.191.69.201     UH          0        0      -  ppp0
 > 204.191.69.201     127.0.0.1          UH          1        0      -  lo0
 > 

------------------------------------------------------------
Rex McMaster                            rmcm@compsoft.com.au 
                                   rex@mcmaster.wattle.id.au
     PGP Public key: http://www.compsoft.com.au/~rmcm/pgp-pk