Not at all, and that was my point. It entirely depends on what you have
going across those PVCs. If we were to attempt to terminate 90 of our
other PVCs running IP on to a single T-1, it would melt immediately. It
would really only take about 12 or so to keep a T-1 very busy.
However, on the specific PVCs you mentioned, I stated that we are not
running a routing protocol and the only traffic is very light SNA stuff.
Even with all 90 PVCs up and running, that interface barely knows that
it's working most of the time. Take a look:
<snip>
5 minute input rate 46000 bits/sec, 86 packets/sec
5 minute output rate 3000 bits/sec, 3 packets/sec
<snip>
That's lower than usual, but as you can see there just isn't that much
traffic and we can quite easily add more PVCs. Just to be on the safe
side, though, we will be adding an additional T-1 later.
>>> "Luong, David" <[EMAIL PROTECTED]> 2/28/01 10:15:29 AM >>>
Hello,
Isn' that an killing the hq T1 circuit if you have 90 PVC terminating
on one
T1?
David.
-----Original Message-----
From: John Neiberger [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, February 28, 2001 8:04 AM
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Re: circuit overbooked?
To answer this intelligently we'd have to get some more information,
such as:
1. What is the average/typical load per PVC right now?
2. What are your typical peak loads?
3. Would peak times on one PVC necessarily correspond with the peaks
on another PVC?
4. What protocols are you running on these PVCs? SNA? TCP/IP? IPX?
5. What routing protocols are you using, if any?
6. How much growth do you foresee?
7. Are your applications able to handle variable latency and delay?
8. Can they recover well if retransmissions are necessary?
Those are the types of things I would consider when oversubscribing a
circuit. With that said, I'll mention two examples in our network.
The bulk of our traffic between branches is TCP/IP and we tend to
terminate between 7 and 10 PVCs on a single T-1. We also have an
additional PVC to each branch that is solely for SNA traffic, which is
very low volume, and we're using static routing here. In this case,
we
have 90 PVCs (coming from 256k and full T-1 circuits) all terminating
at
a single T-1 here at headquarters.
At the moment we're barely pushing that second T-1, but due to growth
needs we will be adding a second SNA T-1 in the future.
HTH,
John
>>> "Jerry Deer" <[EMAIL PROTECTED]> 2/28/01 8:29:16 AM >>>
If i have a 256k host ckt, how many remote 256k ckts with a cir of
64k
can
i have pointing at the host before i have an over utilization problem?
Will
the cir of the host be a key factor?
Thanks for any and all replies,
Jerry
_________________________________
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
