[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: next step
Will Eatherton wrote:
>Once again without offering opinion yet, dont underestimate the complexity
>of trying to standardize the overlal mechanisms of datapath packet
>filtering that are used as basis for wide variety of features. I also would
>be curious what other folks had in mind as to level of specification psamp
>would provide on these sorts of functions.
>
Based on the messages on the list the consensus seems to be that:
1) If we want filters, they should be part of the selector that is in
the line card/whatever device supports psamp
2) Simple filters can be implemented at high speed at small cost so they
won't hinder widespread adoption of psamp
3) They could be very useful for some applications
Just to make sure that we're talking about the same thing - here's what
I think filters could look like:
All packets that go to IP address A
All packets that go to prefix B
All TCP packets that go to port 80 to any address within prefix B
I propose that:
1) We explicitly include in the chapter that the psamp wg wants to deal
with these filters
2) The drafts we produce very clearly and unambiguously define what
filters are
I recommend changing point 1 in the chapter document this way (addition
in CAPS):
1. Selectors for packet sampling. Specify a set of primitive packet
sampling operations for network elements, and the ways in which they can
be combined. SPECIFY THE TYPE OF FILTERING OPERATIONS THAT SELECTORS CAN
PERFORM. This set shall be sufficiently rich so as to support some
existing and emerging packet sampling schemes, and some packet
measurement requirements of some other IETF WG's.
It might make sense to make filters (or more complicated filters (e.g.
the ones that involve digging up fields from the TCP header which is not
at a fixed location because of IP options)) optional, but I think it's
best to address these details later.
Cheers,
Cristian
--
to unsubscribe send a message to psamp-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/psamp/>