Re: [Idr] TCP & BGP: Some don't send terminate BGP when holdtimer expired, because TCP recv window is 0

[Robert Raszuk <[e-mail]>]

Good catch Enke !

Also what if TCP rcv takes the BGP messages and passes it to BGP I/O InQ
which drops it for some reason right there ? Looks to me like we are not
going to detect any event like this here. But the problem we are trying to
address will persist. I think in this thread we are focusing too much on
transport vs application level detection.

And I will repeat the question already stated ... Why rcv would not close
the session in spite of missing KEEPALIVES or UPDATES ?

Tx,
R.

PS. Side note: BGP Operational Message addresses this type of
inconsistencies by periodically comparing BGP Adj_RIB_In and _Out counters.


On Thu, Dec 17, 2020 at 3:41 AM Enke Chen <[e-mail]>
wrote:

> Hi, Folks:
>
> Regarding the patch for openBGPD pointed out by Job, I do not think it
> would work. When the TCP rcv window from the remote is 0, the BGP keepalive
> can still be queued to the socket buffer. It can take a long time for the
> socket buffer to be filled up by BGP keepalives.
>
> It seems that the TCP_USER_TIMEOUT option can be used for the persistent
> zero-size window issue.  The timeout value could be multiples of the
> holdtimer (with min and max adjustments), perhaps somewhere around 5 or 6
> minutes.
>
> Thanks.   -- Enke
>
> ----------
>
> Job Snijders <[e-mail]> Tue, 15 December 2020 21:54 UTC
> <https://mailarchive.ietf.org/arch/browse/idr/#>
>
> [snip]
> How to solve this? Claudio Jeker took a look at what it would take in
> OpenBGPD and came up with the (tiny!) following patch, should be
> readable to most: https://marc.info/?l=openbsd-tech&m=160796802508185&w=2
>
> _______________________________________________
> Idr mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/idr
>