SUMMARY - Network Problems

Sorry for the late summary, but we were trying to solve the problems first
:)

Thanks to :
Figaro, Nicolas
Ralf Van Dooren
Bob Cummings
Praveen P Nair
Thomas M. Payerle

And Special THANKS to Karyn Williams, who really help me to get this problem
sorted.

The problems was with the reverse lookup tables. The servers are in
differents DNS zones, and to get the other reverse lookup tables they need
to contact a higher DNS server, which doesn't have the reverse lookup
configure properly to answer with the right name.

The workaround to get our servers talking properly, was to add the reverse
lookups on each of our server:

Add to your named.conf:

 zone "10.in-addr.arpa" in {
        type master;
        file "master/10.rev";
};

add to wherever you point the conf file to:

# cat 10.rev
$TTL    86400
@       IN      SOA     zeus.rgu.ac.uk.  root.rgu.ac.uk. (
        2       ;Serial
        10800   ;Refresh
        3600    ;Retry
        604800  ;Expire
        86400 ) ;TTL

        IN NS   zeus.rgu.ac.uk.


32.1.1  IN      PTR     zeus.rgu.ac.uk.
42.1.1  IN      PTR     galahad.rgu.ac.uk.
52.1.1  IN      PTR     galahad.rgu.ac.uk.

*       IN      PTR     rfc1918-rgu.ac.uk.

Of course you will have to fix the data so it is correct, but this is the
basic file layout. Mailbox.rgu.ac.uk has the named.conf set up and some kind
of zone file, but none of your data in it. See below. I would ask them to
add it.


My original post was:

Hi gurus,
	First the situation:

We have three vlans configured on the network, the three of them are
connected trough a CISCO core system (some kind of huge router/switch super
smart device, that I don't have access to). In all the vlans are some Unix
server (Solaris) and a lot of window$ XP machines. BTW the same network is
running Novell 6 too. Now in vlan1 I have a Solaris 9 V480 server running
and talking with everyone in the same vlan quick and without problems. In
vlan2 and vlan3 there are some Solaris 7 and 8 (on E450, E220, sparcstation5
and SPARCstation) running on the same way, smoothly connecting with the
other servers and the clients. The problem is when we try to put all the
vlan to talk among them... My server on vlan1 take like 6 minutes (from my
watch) to telnet to any server on vlan2 or vlan3, when you get connected
everything work fine. You can see the same behaviour in many other apps, for
example, a quick test for network connectivity is ping, when I run ping from
my server on vlan1 to vlan2 or vlan3 it takes 30 seconds for the second
package to go trough (yes, the second, the first on goes perfect, but the
second takes time come back), but if I do the ping from the vlan2 to my
vlan1 it takes like 6 minutes for the second package. OK, so we put a
sniffer at the exit of vlan2 server and ping vlan1, same behaviour on the
server, but the sniffer see all the packages coming and going without any
problems/delays.
Now the weirdest, if we do a ping from a pc from the same vlan (even from
the same port on the switch) to the server, it works perfect!

I'm out of ideas here... Anyone could give me anything to test/try?

Thanks.-

Pablo.-
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers