Re: maximum ipv4 bgp prefix length of /24 ?

[Jon Lewis <jlewis () lewis org>]

On Sun, 1 Oct 2023, Matthew Petach wrote:

> If your device is counting on FIB compression to deliver sufficient space savings to allow a FIB of size > SRAM to fit 
> into SRAM, 
> it really should have a reasonable, sane fallback mode for when the next routing update happens to result in a FIB that 
> is incompressible.
>
> Unfortunately, many coders today have not read Godel, Escher, Bach: An Eternal Golden Braid, 
> and like the unfortunate Crab, consider their FIB compression algorithms to be unbreakable[0].

IMO, fib compression is a bandaid that allows putting off forklift 
upgrades.  My personal experience with it has been that it [at least on 
Arista gear] works well (reducing the v4 table from ~925k routes to under 
600k for devices with several full views and several dozen eBGP peers, 
even greater savings for internal devices that only see a full table of 
best paths from a couple of edge devices, and for v6, the savings are even 
greater).

On Arista, I assume the failure mode will be the same as what happens when 
Lpm is exhausted without fib compression.  Routes that don't fit are not 
programmed from the RIB to FIB, and bits of the Internet cease to be 
reachable.  At that point, You'd have to simply reduce the number of 
routes accepted and hope you can get hardware upgrades done real soon.

Anyone running Arista gear with Jericho/Jericho+ chips and full routes, 
AFAIK, must be using fib compression at this point, and probably has at 
least a couple more years to figure out their upgrade path.

----------------------------------------------------------------------
 Jon Lewis, MCP :)           |  I route
 StackPath, Sr. Neteng       |  therefore you are
_________ http://www.lewis.org/~jlewis/pgp for PGP public key_________