All,
On 05/01/2007, at 8:26 PM, Shin Yamasaki wrote:
From: Jas Webb <jaswbb@yahoo.com.au>
To: wg-rms@lists.apnic.net
Date: Fri Jan 05 2007 09:55:19 GMT+0900
Perhaps the WG could ask Terry (and APNIC) to re-take
a position as "WG advisor" or similar that may
maintain the impartiality but also be the 'buy in' and
point of responsibility from the secretariat.
In my opinion, we don't have to re-appoint Terry special position
officially. The reason is Terry has already been providing enough input
to this WG. In other words, he already takes that kind of role you
mentioned. I prefer to have more participants rather than having more
officers in this WG.
I tend to agree with Shin. I am a member of the WG with a special
interest in seeing that the participants are furnished with all suitable
information - The secretariat is equally interested in the proceedings
of this WG as it could revolutionise the member interaction.
Further, it seems that this working group has a longer
term ongoing role that actually encompasses the
DB-SIG, I think this WG should be promoted to a SIG
level replacing and consuming the DB-SIG, which I
don't see any traffic on.
Since this WG is a child of Policy WG, I don't think it's appropriate to
take another SIG over. You can propose reform or re-org the DB SIG
after this WG is closed.
I have redirected this suggestion to the policy people at APNIC and will
await their response before commenting further.
well.. sort of.. I think way to aggressive, in the
same phrase as "I don't want to spend money on having
someone deal with email update issues", I equally
don't want have to rapidly change a resource budget to
pay for a rapid implementation.
how about this, After adoption:
6 months: Start accepting all objects via new method
10 months: stop accepting domain objects via email
14 months: stop accepting inetnum, inet6num, autnum
objects
18 months: stop accepting all remaining objects.
Any comments, especially from APNIC secretariat?
There is an underlying argument in when we start and stop accepting
updates from different entry points. This argument centres around the
question: "which is the authoritative data set?".
At present, due to facets of history, APNIC has two distinct databases,
one known as registry and one known as whois. Which do we consider to be
the true and correct representation? After evaluations of workflows,
data sets, and processes it was agreed internally that the 'registry'
data should be known as the one and true location of information. Whois
would then be a view of registry that fits all of the APNIC community
and business rules (such as reporting, privacy and confidentiality).
The XML mechanism proposed updates 'registry' directly and conversely
email updates 'whois' alone. If we were to run both mechanisms for a
data-set, say inetnum, we could have a clash of data and that raises
some very difficult data-merge questions when you have two equally
authorised data-sets.
Based on this, and I take your point about the timing of the change. Can
I suggest:
6 months: Have prototype for all objects available
10 months: stop accepting domain objects via email
14 months: stop accepting inetnum, inet6num, autnum
objects
18 months: stop accepting all remaining objects.
Is it possible for APNIC to have a working prototype 2
months after adoption so that we can begin change-over
work?
APNIC must also provide tools for this.. perl or java
please.
They might already have that.
We have perl tools available, you may download these from:
ftp://ftp.apnic.net/pub/utils/db-tools/
I will pass the Java request to the APNIC software team.
From my opinion - this reads as something deeper. Why
the need for aggressive implementation? Perhaps Terry
knows more than is written in the proposal.. I looked
back over past APNIC meeting agendas and Terry has
previously presented about issues with DNS generation
and the APNIC internal registry system. It seems
plain that APNIC has a system that is desperate for an
overhaul, and this proposal is that overhaul. If that
is the case, and the proposal is about replacing an
old incapable system, then I'll double my support of
the proposal (with modifications).
I leave this for Terry.
The existing registry system is under constant development to improve
the processes and workflows so that APNIC can provide better and faster
responses to members and NIRs alike. These continual improvements have
gone some distance, however we believe we have reached a fork in the
road when it comes to what a split whois/registry system can provide the
membership.
I feel that by deprecating email we not only resolve some long standing
issues, we also remove a capability ceiling that the existing system
has. Certainly I have previously presented the failings of various
aspects of the registry to our membership. The failings, for interest
sake, were created by data-merge complexities formed from parallel but
different data sets.
Jas, I think 'overhaul' is a strong word, and creates visions of a
complete replacement, that is not really required. What is required is
continuous improvement and looking at the suitability (in this case the
unsuitability) of email updates is the next logical step to a long-term
member focused registry system.
Terry
--
Terry Manderson email: terry@apnic.net
Network Operations Manager, APNIC sip: info@voip.apnic.net
http://www.apnic.net phone: +61 7 3858 3100
_______________________________________________
Wg-rms mailing list
Wg-rms@lists.apnic.net
http://mailman.apnic.net/mailman/listinfo/wg-rms
