[c-nsp] ATM IMA Overhead....

[Richmond, Jeff (ELI)]

Regarding IMA specifically, your typical IMA overhead is very small and
depends on the Frame Length. The average ranges from .4% to 3.2% (32, 64,
128 and 256 - not sure what values Cisco supports, but 128 is the most
common, and that is just under 1%). The smaller the frame Length, the
quicker the response to changes, but the more overhead, and vice versa. This
is in addition to the normal ATM overhead, so keep that in mind.

regards,
-Jeff

-----Original Message-----
From: Dan Hollis [mailto:goemon at anime.net]
Sent: Thursday, August 05, 2004 2:30 PM
To: THOMAS, MEAGAN (PBI)
Cc: 'cisco-nsp at puck.nether.net'
Subject: Re: [c-nsp] ATM IMA Overhead....


On Thu, 5 Aug 2004, THOMAS, MEAGAN (PBI) wrote:
> Hey, this may be off base, but what the hey, I gotta try....
> So, does anyone have any resources that can direct us to the kind of
> overhead ATM can cause when using IMA?  Cisco's IMA docs suggest 10-12%,
but
> ATM overhead varies - other documents on Cisco's site suggest ATM overhead
> is like 20%, depending on packet size.

The biggest impact appears to be the encapsulation. I did tests a year or 
two ago which indicated aal5snap overhead was around 22% whereas aal5mux 
is around 13%. This was with mtu 1500.

-Dan


_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/