[c-nsp] ASR 1k vs 9k as a non-transit BGP router with full tables?

[Patrick M. Hausen]

Hi!

> As for comparisions - 1001/1002/1002F are no longer in game,
> and while they perform decently for control plane, even first gen ASR 9k’s
> (like 9001/9001S and RP1s for 9006/9010) will beat them unless you
> go for RR role in a specific config (SRD) - as Ytti mentioned, BGP on
> 32 bit IOS-XR has memory limitations per process. This is not
> Patrick use case however I believe.

The use case is simply "full tables BGP" with currently 4x 1GB/s
uplinks and possibly 6 in the near future. Upgrade to something
10G-ish not planned at the moment. 300-400 Mbit/s aggregate
traffic across all uplinks currently.

So we are too memory heavy for the C6500 (SUP720-10G) and
then there's the TCAM limitation ... although our bandwidth requirements
are rather small. And then the C6500 definitely starts to rot - I wonder
if I will ever get anything beyond 12.2(33)SXJ10 if (when!) the next
remote security bug hits.

> Going back to original question - if that’s going to be refubrished unit,
> 9001 propably fits the bill in the best way. 9904 in the new data center
> is propably the best choice given your requirements.

Yeah, the 9904 looks quite nice. Don't know if it's available refurbished,
already. The 9001 would probably cost us 3 to 4 times as much per
box as the 1001-X. I haven't received a written quote yet and I'm
unsure about the cost if the 20G and 2x10GE licenses. We'll see.

> Stay off the ASR9k cluster licenses BTW :) You don’t need them for
> your use case.

So I figured already. The new boxes will be all layer 3, so no need to
mess with VSS and sons to get multi-chassis portchannel etc.

Thanks to all for your valuable input.
I'll report what we got ;-)

Patrick
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 496 bytes
Desc: Message signed with OpenPGP
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20170803/cf9f1882/attachment.sig>