Re: [IPFIX] IE-doctors: transitioning IEs
Paul Aitken <[email protected]>
| Newsgroups | gmane.ietf.ipfix |
|---|---|
| Message-ID | <[email protected]> |
Brian, What is the conclusion to this? ie, do you want to rework ie-doctors, and I'll polish and publish my -00 for IE equivalence? P. On 19/07/12 13:45, Brian Trammell wrote: > Hi, Paul, > > I'm not much of a fan of sticking enterprise-specific references in the registry, actually. It was suggested that this is a problem (indeed, I believe it is), and this is indeed a solution to it -- but it seems like a much cleaner mechanism for doing this would be to allow the actual devices doing the exporting to annotate the output data with an IE map (much like type information export as in RFC 5610). > > One clarifying point: the Collecting Process, of course, needs to be updated too, in order to understand the Options providing the IE map. But this only happens once, as opposed to the registry-based solution which requires the CP to be regularly updated (or to automatically query) the IANA registry. > > Removing a whole column and subset of sections from the ie-doctors at this late date may be a little problematic procedurally; I'm not sure of the right thing to do, and would turn to our chairs and AD for guidance... > > Cheers, > > Brian > > > On Jul 19, 2012, at 2:16 PM, Paul Aitken wrote: > >> Brian, >> >> In my IE-doctors reviews, I expressed some reservation about recording Enterprise-Specific information in IANA's IPFIX registry as specified in this text: >> >> In order to support transition from experimental registration to IANA >> registration, the IANA registry provides an optional "enterprise- >> specific IE reference" column for each Information Element. In cases >> of promoted enterprise-specific Information Elements, this column in >> the registry SHOULD contain the private enterprise and Information >> Element numbers of the enterprise-specific version of the Information >> Element. >> >> >> I've had a similar, but automated, idea in the back of my mind to for some years: an upgraded device could export a table mapping oldIE to newIE, where either of these IEs can be IANA standard or Enterprise Specific. >> >> The table would be interpreted as "(if) I used to export oldIE, now I export newIE", so the Collecting Process can readily re-map exported data over the Exporting Process software change boundary. If the Collecting Process never received any oldIE elements, it can disregard the information. >> >> With the IE doctors / IANA registry idea, the information has to be in the registry and the Collecting Process has to have read the new registry. So both the Exporting Process and Collecting Process must be updated. >> >> With my idea, only the Exporting Process is updated, and it informs the Collecting Process. >> >> I drafted a -00 on this last year which requires a little more work, so I've not published it yet. >> >> Of course the two mechanisms are complementary rather than mutually exclusive. It's a question of which issues we're trying to solve. >> >> P. >> >> _______________________________________________ IPFIX mailing list [email protected] https://www.ietf.org/mailman/listinfo/ipfix