01-08-2013 02:34 PM - edited 03-07-2019 10:58 AM
I'm looking at the possibility of purchasing a few 3750 to build a switch stack to replace my current core and access switches. I'd like for the stack to support server, SAN, and user traffic, as well as provide simple switch management, redundancy, and failover. I've also been looking at the 4500, 6500, and Nexus series switches (although the Nexus switches maybe a bit too much for my network).
My current environment consists of 3560's as the core/distribution/access switches, a mixture of Hyper-V and VMWare environments (although we're slowly migrating to Hyper-V), an HP SAN (P4500), and about 100 users.
Has anyone faced any issues going with the 3750's or have an alternative, more efficient solution?
01-08-2013 02:47 PM
If Nexus (which is designed for Data Centre networking) is too much for you, then 4500R+E with Sup7E or Sup8E will be a good fit for you.
Before you go off getting one, how many 10 Gbps links will you be using?
01-08-2013 02:55 PM
I agree I am so not a fan of 3750's and the whole stacking idea. With 4500E and X you can use VSS which I prefer:
http://www.cisco.com/en/US/prod/collateral/switches/ps10902/ps12332/at_a_glance_c45-696645.pdf
01-08-2013 03:14 PM
What is the requirement? If you need unified ports (FC, FCOE, Ethernet) the Nexus 5K/2K combo is good choice. Also, as Leo noted, how many 10Gig and 1Gig ports do you need? How about copper vs fiber? If you are moving from 3750s to 4500 or 6500, you are talking 110 vs 208, so big power change. Also, rack space and cooling.
HTH
01-09-2013 06:08 AM
We do not utilize FC or FCoE, so unified ports are not a requirement. I'd estimate we need about 265 1Gig ports, and very few 10Gig ports.
We have an ample amount of rack space and proper cooling, so I don't think those will be an issue.
I've also heard that the 4500's are not usually used at the core layer, since it can be easily oversubscribed. Any truth in that?
01-09-2013 06:36 AM
In this case, for your core/distro a set of 6500 with Sup-2T and VSS enabled should do the job. With VSS, there is no need for HRSP, VRRP, etc...It also eliminates the need for spanning tree since 2 6500 logically become one device, but you still run it just in case there is a split brain.
HTH
Good Luck
01-09-2013 06:55 AM
Would I be able to configure this stack to act as an access switch as well, making this an all-in-one core/distro/access layer switch?
01-09-2013 04:44 PM
Yes, you can use the 6500s as access/distro and core switches, but you don't need to make them VSS as VSS is designed for distro/core or data centers where you can dual uplink your access switches to it or dual connections from servers. When you connect a server using two links to a VSS pair, you have NIC/link redundancy and also switch redundancy. You can however use the 2 switches as 2 different units and run HSRP, VRRP, etc... and that should work fine.
Good doc to read.
http://www.cisco.com/en/US/docs/solutions/Enterprise/Campus/VSS30dg/VSS-dg_ch1.html
HTH
01-08-2013 07:19 PM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
The advantage of the 3750 series, small stacks are generally less expensive than chassis solutions, especially when providing device redundancy.
For a large number of ports, chassis may be less expensive. Depending on the architecture of the chassis, it may offer better raw performance too. For example, 4500s now provides up to 48 Gbps (fabric) per slot, 6500s now provide up to 80 Gbps (fabric) per slot. 3750s have up to 32 Gbps (dumb ring) per stack, 3750-E/X have up to 64 Gbps (not-as-dumb ring) per stack member.
Chassis (or high end stand alone switches) generally have more hardware resources and sometimes additional features. IPv6 can be more problematic on the 3750 series as it consumes hardware resources (particularly the TCAM).
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide