Hi Eric,
I can add the text "and MUST drop the packet" to section 9. I think it’s a MUST
and not a SHOULD as effectively the SFF cannot forward the packet if the SR
forwarding state is missing from the cached. So the new text will be:
"In the case a packet arriving from an SF does not have a matching
cached entry, the SFF SHOULD log this event, and MUST drop the packet."
For section 5.2 I am not sure it is correct to empty the cache as this would
cause the SFF to populate/flush the cache for each packet in a particular
traffic flow. It seems more appropriate to base the cache flush on some timer
and this is already specified in section 9:
The cache mechanism must remove cached entries at an appropriate time
determined by the implementation. Further, an implementation MAY
allow network operators to set the said time value.
Jim
-----Original Message-----
From: Joel Halpern <j...@joelhalpern.com>
Sent: Monday, April 24, 2023 9:48 AM
To: Éric Vyncke <evyn...@cisco.com>; The IESG <i...@ietf.org>
Cc: draft-ietf-spring-nsh...@ietf.org; spring@ietf.org;
bruno.decra...@orange.com
Subject: Re: [spring] Éric Vyncke's No Objection on
draft-ietf-spring-nsh-sr-12: (with COMMENT)
Thanks Eric. Trimmed, just to keep your two questions. I am speaking as a
significant contributor, but will leave text changes to the pen holders.
Yours,
Joel
On 4/24/2023 9:12 AM, Éric Vyncke via Datatracker wrote:
> Éric Vyncke has entered the following ballot position for
> draft-ietf-spring-nsh-sr-12: No Objection
>
> ## Section 5.2
>
> Should the cache be emptied when a packet came back ? or on time-out
> (as either no reply or multiple replies could be expected). The reader
> has to wait until section 9 to get some hints (and it is unclear
> whether the caching refers to the section 5.2 caching).
Yes, the cache should be emptied in both cases. And it would probably help
readers to say so earlier in the document.
>
> ## Section 9
>
> In absence of cache entry, should the text specify that the packet is dropped
> ?
> Or should a ICMP sent back to the SF ?
The packet should be dropped in the absence of a cache entry. I see no point
in sending an error to the SF, as it would likely have no idea what to do with
the message. Agreed that we should probably say so.
>
_______________________________________________
spring mailing list
spring@ietf.org
https://www.ietf.org/mailman/listinfo/spring
