Hi George, George Michaelson wrote:
Hi. Sorry for comparative silence from APNIC, we're back looking at getting the V3 code beyond test-compile state into pilot service.
Andy Linton has been looking at this with me and we're now pretty sure the only schema element we don't have in APNIC 181 objects in-hand is the referred-by field. Having worked out its a nic-hdl, we figure that for 'old' data we just make this the existing meta-maintainer for APNIC and can do that easily in ripe2rpsl.
If you mean referral-by, it points to mntner handle, mandatory in mntner object and is part of RPSS (RFC2725) spec.
But we have this problem about having country: fields in domain/inetnum/person objects. If we modify schema here, re-compile whoisd, are we able to run NRTM feeds to you and have your side strip non-conformant attributes, or is there some relax mode for NRTM where it doesn't care?
The core update server (whois_rip, and NRTM is part of it) cares only about primary keys. It also parses (so they should be RPSL compliant) and stores attributes that are part of lookup/inverse keys. However, they are considered as optional by NRTM module. Most of the syntax checks are done by dbupdate. That's where you need to modify definitions. But as far as mirroring is concerned, this is not an issue.
My reading of code is a bit hazy here. I can't see where you are going to bypass schema checks in NRTM. If thats the case, we'd have to either overload remarks: fields to preserve this data (which is yucky) or you need to revise the RIPE schema.
We should be able to take a feed incoming anyway (and would like to do that asap) but we're worried about how you can take a feed back from us if we do the codechange in the defs/* here.
cheers -George -- George Michaelson | APNIC Email: ggm@apnic.net | PO Box 2131 Milton QLD 4064 Phone: +61 7 3367 0490 | Australia Fax: +61 7 3367 0482 | http://www.apnic.net
Regards, Andrei -- Andrei Robachevsky DB Group Manager RIPE NCC