[c-nsp] Re: Cisco 7600 vs Juniper M7i
Church, Chuck
cchurch at netcogov.com
Thu Jun 2 13:37:42 EDT 2005
Thanks, Ian. I was thinking more about the Flexwans, which are really a
close relative of the VIP, right? I haven't had any experience with the
OSMs. Those use the Sup's PFC for forwarding, even if source and dest
is 2 different ints/subints on the same OSM? If so, I retract my
previous statement! That would seem to address the 'VIP running at 99%'
issue we sometimes see under DOS'es. The new SIP-200 had a PPS rating
associated with it, I'm assuming that one doesn't use the PFC?
Chuck Church
Lead Design Engineer
CCIE #8776, MCNE, MCSE
Netco Government Services - Design & Implementation
1210 N. Parker Rd.
Greenville, SC 29609
Home office: 864-335-9473
Cell: 703-819-3495
cchurch at netcogov.com
PGP key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x4371A48D
-----Original Message-----
From: Ian Cox [mailto:icox at cisco.com]
Sent: Thursday, June 02, 2005 12:56 PM
To: Simon Leinen; Church, Chuck
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] Re: Cisco 7600 vs Juniper M7i
At 04:54 PM 6/2/2005 +0200, Simon Leinen wrote:
>Church, Chuck writes:
> > Is this really a fair fight though? You're comparing a router
> > against a high-end layer 3 switch on just Ethernet. Everything they
> > tested was things the 7600 can do on the PFC.
>
>Note that PFC3B/PFC3B-XL can do many things (IPv4/IPv6 uni/multicast,
>MPLS, various forms of tunnel encapsulations, IPv4/IPv6 Netflow (they
>can only export IPv4 flows for now though), many kinds of QoS)...
>
> > I'd expect the Cisco to destroy the Juniper on that. Now if they'd
> > tested OC-x interfaces, that'd have been a totally different story,
> > I bet.
>
>We have a few OC-12c and OC-48c POS interfaces (OSMs) on the 7600, and
>they perform fine. They are quite expensive though, the port density
>isn't that impressive, and they cannot use the Sup720 switching fabric
>(but the 32 Gbps legacy bus supports quite a few of them). Forwarding
>is done by the PFC even for these interfaces.
The OSMs do use the Sup720 switch fabric to move packets between line
cards, running at 8Gbps. They use the bus just to pass the packet
headers to the PFCx to make the forwarding decision.
wodonga#sh mod
Mod Ports Card Type Model
Serial No.
--- ----- -------------------------------------- ------------------
-----------
1 4 2-port CHOC-12/DS0
SI OSM-2CHOC12/T1-SI SAD0625024U
2 24 CEF720 24 port 1000mb
SFP WS-X6724-SFP SAL08394EXH
3 4 2-port OC-12c ATM MM+ OSM-2OC12-ATM-MM
4 0 4-subslot SPA Interface
Processor-200 7600-SIP-200 JAB080504XZ
5 2 Supervisor Engine 720
(Active) WS-SUP720-BASE SAD07170004
7 4 2-port OC-48c POS
SM OSM-2OC48/1DPT-SS SAD060202HU
8 4 1-port OC-48c POS
SM OSM-1OC48-POS-SS SAD0537012R
9 16 SFM-capable 16 port 1000mb
GBIC WS-X6516-GBIC SAD044908G9
wodonga#sh fabric status all
slot channel speed module fabric
status status
1 0 8G OK OK
2 0 20G OK OK
3 0 8G OK OK
4 0 20G OK OK
5 0 20G OK OK
7 0 8G OK OK
8 0 8G OK OK
9 0 8G OK OK
>The SPA/SIP modules that Simon mentioned should improve choice of
>ports, port density, and performance (I think these carrier cards *do*
>have fabric attachments).
The Enhanced FlexWAN, all OSMs, 7600-SIP-200, 7600-SIP-400 all have
fabric connections.
Ian
>--
>Simon.
>
>_______________________________________________
>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/
More information about the cisco-nsp
mailing list