Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
(ASCEND) Multiple MAXen w/o OSPF?
Just out of curiosity, is there anyone out there using more than 1 MAX
on a single LAN segment (sharing the work load with multiple PRI's) and
NOT using OSPF (got to be several I'm sure)? I have a situation where I
need to utilize two MAXen side by side running Appletalk and IP (thus
the MAX cannot run OSPF), but I also need to connect it to a Cisco
running OSPF to the rest of our network.
We currently are using RIPV2 on both MAXen and the Cisco which is kind
of...well, limping along. We've been experiencing connection delays (up
to 30 seconds) with user IP sessions due to the RIPV2 updates and router
suppressed hold times, so we came up with a scenario to override the
problem.
The Cisco (router A) is directly connected to both MAXen (routers B and
C). In Cisco (router A), we added static route entry's (with a higher
metric than default) for all subnets related to the dialup clients, and
pointed it to MAX (router b). In MAX (router b), we added the same
static routes (again, with a higher metric than default) to point to MAX
(router C). Anyone dialing into MAX (router C) gets connected and
immediately gets an IP session (via default route from A to B to C,
until RIPV2 kicks in to optimize a direct connection from A to C). If
he losses his connection and then reestablishes within a minute or two,
the static entry's kick in (overriding the RIPV2 hold down timers) and
the user gets his packets (via A to B to C). The hold down timers
expire after ~180 seconds at which time he then gets his packets
directly from Cisco (router A).
Now, if someone dials into MAX (router B), the session is established as
before and everything appears to work. If he looses his connection and
then reestablishes within the 180 seconds, the static route kicks in and
he gets his packets directly from the Cisco (router A). The hold down
timer expires, at which time he then the gets his packets (again)
directly from the Cisco (router A) from the RIPV2 update.
We are now seeing the delay times for all profiles down to a minimum
(acceptable for the moment), but now the performance (transfer rates)
for anyone connected to MAX (router B) has dropped ~20-30%. As far as I
can tell, the routing is correct (theoretically) but it appears the
MAXen together are doing something with MAX B's packets. The Cisco
looks happy with its configuration and is sending the packets to either
MAX router (depending on which MAX the call was placed). Can anyone
confirm this?
Now, I'm no routing expert but this configuration is anything but
optimal for more than several reasons. So, what I'm looking for is some
advice from those out there with similar configurations and how you got
around this issue.
TIA
Kirk
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>