true- cef if the best for most situations, certainly at the 7500 and gsr
levels where mlppp is a joke.. i m just saying for 2600 with 2 t's, also i
have experienced lots of cef problems with NAT, which you would normally do
on a little 2600. mlppp for me has not had these issues.. but i agree if y
> Yes I would use mlppp and ios in the same sentence, as I have
> been running it without a hitch on a 7200 for 6 months. also,
> you guys are missing his point (t-1's to the internet) what ISP
> is going to run
> OSPF or EIGRP with a customer ? please.
I don't see anything in the original post a
""Brunner Joseph"" wrote in message
[EMAIL PROTECTED]">news:[EMAIL PROTECTED]...
> Yes I would use mlppp and ios in the same sentence, as I have been running
> it without a hitch on a 7200 for 6 months. also, you guys are missing his
> point (t-1's to the internet) what ISP is going to run
> OSPF
Yes I would use mlppp and ios in the same sentence, as I have been running
it without a hitch on a 7200 for 6 months. also, you guys are missing his
point (t-1's to the internet) what ISP is going to run
OSPF or EIGRP with a customer ? please.
Maybe, if they managed the router, but he didnt say t
I didn't say that all MPPP bundles were flaky. The majority are not. But
on the whole, it seems that there is a greater chance of flakiness within
bundles than in CEF implementations.
When I say flakiness, it generally boils down to what you had said before -
lots of overhead and buffering inv
9/2002 09:04 AM
Please respond to "nrf"
To: [EMAIL PROTECTED]
cc:
Subject: Re: Combining T1's into one pipe [7:46942]
Uh, really? You sure about that.
>From my experience, when you're talking about IOS, you should never ever
use
the term
Okay. perhaps I don't monitor my bundles as close are y'all do, but
we're running quite a few sites using multiple T1s bonded with MLPPP, and we
don't have any stability problems (as far as dropping traffic and EIGRP
neighbor changes, etc) that I'm aware of nrf, tell me more about what
to
If you're running a dynamic routing protocol (i.e. RIP, IGRP, EIGRP, or
OSPF), they should see the two T1s as equal cost paths and automatically do
per-destination load balancing (if you're running CEF, then that can be
per-packet, at least with EIGRP, but I would suspect the same no matter now
th
its definately worth it.. combine multiple pipes at layer 2. I use MLPPP
with my ISP and it rocks.. forget all those shaky stupid CEF
and PER-PACKET configurations.. if you can get PPP going between your
carrier and you, you can get it all going to one router on their side, then
you should run MLP
Uh, really? You sure about that.
>From my experience, when you're talking about IOS, you should never ever use
the terms MPPP and stable in the same sentence.
I recommend CEF not because it's not flaky, because it is, but because it's
a lot less flaky than Cisco's MPPP implementation.
""Brunn
I have two point to point T1's that I'm thinking about combining (known as
NxT1). Both connections are going through the same routers at each end (4700
and 2600). I found some information on Cisco's website but they mention that
it can be complex and talk about alternatives.
Is anyone else doing
11 matches
Mail list logo
