Re: [IPFIX] IPFIX layer 2 packet count

Brian Trammell <[email protected]> Sat, 19 Apr 2014 09:04:01 +0200

Newsgroups	gmane.ietf.ipfix
Message-ID	<[email protected]>

hi Paul, all,

Given that we did the same thing with transportPacketDeltaCount (i.e. only packets with transport layer content), where (for TCP) transportPacketDeltaCount < packetDeltaCount less bare SYNs/ACKs (or other transport layer control), it makes a lot of sense to have new frame count IEs.

We'd probably want the description to note that this should only be exported by EPs attached to MPs that can actually see layer 2 (i.e., that aren't working only on the IP layer and above)... though on second thought that's a little obvious...

For naming, layer2PacketDeltaCount would be more consistent with other IE naming, but I think frameDeltaCount is a better, more precise name. So I'm good with either actually.

Thanks, cheers,

Brian

On 18 Apr 2014, at 23:50, Paul Aitken <[email protected]> wrote:

> Andrew,
> 
>> Is there a situation where a hypothetical flow with both packetDeltaCount(2) and frameDeltaCount(#TBD) would have different values?
> 
> If any of the observed layer 2 frames don't contain layer 3 packets, frameDeltaCount would be > packetDeltaCount.
> 
> P.
> 
> _______________________________________________
> IPFIX mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/ipfix

_______________________________________________
IPFIX mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipfix

signature.asc (application/pgp-signature, 496 B)