Cisco 3750-x Routing performance

[hobbe]

Hi

I am trying to findout what the routing performance of a 3750-x is

The documents clearly states

High-Performance IP Routing

Cisco Express Forwarding hardware routing architecture delivers extremely high-performance IP routing in the Cisco Catalyst 3750-X and 3560-X Series Switches.

But what does that mean when it comes down to numbers ?

"extremely high-performance" is not explaining a whole lot.

any help that can point me in the right direction is greatly apreciated

[johnnylingo · Accepted solution]

There's 3 separate numbers to look at when dicussing routing performance:

1) Max packets per second

2) Max throughput

3) Max number of routes

#1 is pretty easy to find, and will greatly influence #2.  However, packet size also comes in to play.  Some will want to do the calculations with a very small packet size (64 bytes), while others like to assume something in the middle like 512 or the max normal size which is 1500.

#3 is usually dependent on platform and in some cases the license level.  Switches are usually limited to a few thousand routes, so using them as Internet routers running BGP is a problem.

You should be able to find these on the data sheet - http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps6406/data_sheet_c78-584733.html

[hobbe]

Hi

Yes I agree that it depends on the number of pps and the packet size. But these are the values that I am looking for and I am for some reason unable to find them.  i can find them for L2 but not for L3.

I can find the max number of throughput for the interface in pps, but that is L2 and how does the L3 compare ?

Are they the same ?

Sorry this paper/link does only state "extremely high performance" and so on.

[Joseph W. Doherty · Accepted solution]

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

Sure, look at: http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps6406/data_sheet_c78-584733.html 's table 9.

Suppose we consider near line-rate or wire-speed capacity, for all ports, "extremely high performance", what would this require?

Well on a 48 gig port model with two 10g uplinks and with two 16 Gbps stack ports, we need 100 Gbps of bandwidth.  If the ports are duplex, we need to double that, so we need 200 Gbps.  All models have 160 Gbps fabric, so we're a bit short for the 48 port models using dual 10gig, but if we don't use the uplink ports, 160 Gbps is just enough for the 48 gig ports and two stack ports.  Or if we only use 4 gig uplinks, we're only a 8 Gbps shy.  So, you might consider this "extremely high performance".

For the same 48 gig ports with two 10g uplinks and two 16 Gbps stack ports, we need 148.8 Mpps for minimum size Ethernet packets.  Again we're shy as the 48 port models are listed as having 101.2 Mbps, only enough for 68 Gbps.  However, as that's exactly what's needed for the 48 Gbps ports and with two 10g uplinks, perhaps stack ports aren't counted as part of device's packet forwarding capacity.  In either case, as real world packets generally aren't all minimum size, there's likely sufficient PPS to handle all ports with "normal" size packets.  So, again, you might consider this "extremely high performance".

[hobbe]

Hi

The link does not show the routing capabilities in pps.

As it states in the link:

Cisco Express Forwarding hardware routing architecture delivers extremely high-performance IP routing in the Cisco Catalyst 3750-X and 3560-X Series Switches.

To me that would indicate that the CEF hardware routing architecture is not the Switching fabric.

ie the switching fabric can not be based for the calculations.

Am I wrong in assuming so ?

What I am looking for is something like the

http://www.cisco.com/web/partners/downloads/765/tools/quickreference/routerperformance.pdf

but with information relevant for the 3750-x

[Joseph W. Doherty · Accepted solution]

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

Generally, on a L3 switch, L2 and L3 performance are the same.  This holds true, I believe, for the 3750 series.

Regarding a switch performance sheet, like the one you referenced for routers, there's: http://www.cisco.com/web/partners/downloads/765/tools/quickreference/switchperformance.pdf

It's old, like the router performance sheet reference, and doesn't list the 3750-X series, but basically a -X performs the same as an -E.  The major difference is the -X provides StackPower.

[hobbe]

Thanx for the information

and thanx for the link to the switch performance.

This is as always a good information source.