[c-nsp] ASR 9000 Series MPLS Label Limits

adamv0025 at netconsultings.com adamv0025 at netconsultings.com
Fri Mar 31 07:33:57 EDT 2017


> James Bensley
> Sent: Friday, March 31, 2017 10:28 AM
> 
> Just to close this up for anyone else with the same query;
> 
> If we look at the output below from an example ASR9001 there is 99,625
> MPLS labels in LFIB. One can see that roughly double the number of TCAM
> entries are used (196,055). The TCAM also shows 2M entries/records as the
> maximum number of entries possible. Each label is using 2 records in TCAM
> (the "records" are joint ingress and egress
> labels) so it can effectively hold 1M labels entries in the 2M TCAM records. A
> few extra entries inside the LFIB and TCAM are used for some internal "stuff"
> so the number show as used in TCAM isn't exactly double the number of
> labels in the LFIB.
> 
Hey buddy,
Sorry I mean to get back to you on this one, but didn’t have any free cycles back then and eventually forgot about it.

My notes from discussion with Aleksandar:

HW label limit on ASR9K platform:
Information required for HW forwarding is stored in the RAM used by the Network Processor (NP). 
Search memory on Typhoon and Tomahawk is DDR3.

You can see the utilisation of this address space by running:
"sh cef platform resource summary location <location>"

You can find the Typhoon and Tomahawk scale limits in any of the recent BRKARC-2003 slides from Cisco Live
The 1M label space is not divided into local and received. It's all one space.

The hard limits (1M labels, 2M MAC, 4M/2M v4/v6 routes) are per NP. 
However, since the content of respective tables is synchronised between NPs, the same scale limit applies at LC level. 
If selective VRF download is disabled (and it is disabled by default), this is also a per-chassis scale.

Hope this helps

adam

netconsultings.com
::carrier-class solutions for the telecommunications industry::



More information about the cisco-nsp mailing list