[c-nsp] [SUMMARY]: 4900M vs. 4503 for core

[Kevin Hatem]

The 4900m is a robust switch with plenty of BW on the fabric. Port density is not plentiful but.......Using the twinG is a choice - just check on the limitation of use not only with using them on the onboard X2 slots, but also ASIC restrictions.  I know that the SUP6E (the 4900m SUP?) uses stub asics to the fabric and has limitations for combining 1G and 10G on the same asic.

The Juniper and HP boxes that others have suggested are good boxes too.  It appears you have some time to investigate many solutions.  The shortage of the 4900 and other such products are derived as a result of limited component production from Cisco's manufacturing plants (overseas).  But the suggestion that Cisco is pushing other products (Nexus) is plausible.

-nuff said.
-kevin.



-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Jason Gurtz
Sent: Thursday, January 28, 2010 15:34
To: cisco-nsp at puck.nether.net
Subject: [c-nsp] [SUMMARY]: 4900M vs. 4503 for core

> Is there anything glaringly wrong with choosing the 4900M using twin-gig
> based connections to the access layer over the 4503 Sup6 and 46xx line
> cards in our situation?

Thanks all for the replies!  A person also responded privately with the
opinion that most people want Netflow down the road.  Unfortunately, since
Netflow has been removed from the 45xx with the Sup6 it would require 65xx
at $$++.  Squarely in the want vs. need bucket for us

Unfortunately, I left out that that most of the gig uplink connections are
fiber so a 3560G doesn't have enough SFP ports.  I did find the
WS-C3750G-12S-E which looks like the good low-cost option.  On the minuses
side, it's a softswitch, and no 10G uplinks for linking in the server
access switches.  The main downside here is advocating for their
replacement and purchasing strategies around here.  eBay, used equip.,
etc... are pretty much verboten.  Basically, if we buy these now, they'll
be here in 5 years and forklifting the network core could be painful.

Point well taken on the stacking related maintenance downtime issue.  We
plan on doing pure routing and GLBP so thankfully this wouldn't affect us.
This issue will bite us with the server access layer. :(  I'll join the
many who want this problem to go away.

The availability issues with 45xx and 49xx shouldn't be a problem as
4507's are being spec'ed for some access switches and we have until
summertime to do this.  It's interesting though, makes me wonder if it's
just really high demand, or C pushing other platforms.

I discovered the 4928-10G, but the 4900M config comes in cheaper,
apparently due to only needing one 8 port card.  I'm assuming the 2:1
oversubscription is not an issue when running these 10G ports at 1G.  Only
thing is 2000W of power supply vs. 600W.  It does seem silly to do the
twingig thing; if only there was a 20-port sfp halfcard!

Thanks again,

~JasonG
_______________________________________________
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/

This e-mail, including any attachments and response string, may contain proprietary information which is confidential and may be legally privileged. It is for the intended recipient only. If you are not the intended recipient or transmission error has misdirected this e-mail, please notify the author by return e-mail and delete this message and any attachment immediately. If you are not the intended recipient you must not use, disclose, distribute, forward, copy, print or rely on this e-mail in any way except as permitted by the author.