RIPE 51

• content to [email protected]
• format to [email protected]

RIPE 51 - Hotel Krasnapolsky - St-Johns II - 20051013 - 14:00-15:30

Chair: Niall O'Reilly
Scribe: Pierre Baume
Jabber: Scott Donald

Webcast and Feedback Archives:
http://www.ripe.net/ripe/meetings/ripe-51/sessions-archive.html

Niall notes that the DNAME issue mentioned at the last meeting isn't
such a problematic one, according to the people reporting it, as it is
a fair assumption that name servers supporting NAPTR RRs certainly
also support DNAME RRs.

With this caveat, the minutes are approved.

No open issues/actions.


C. Carsten Schiefner - Deutsche Telekom (10 min)
News summary:
. IETF 63
. I-D/RFC watch
. DENIC's ENUM day


The presentation is at:
http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-news.pdf

No question.


D. Axel Pawlik - RIPE NCC (15 min)
Status update of ENUM at the RIPE NCC
. Statistics
. Tier 0 status
. misc


The presentation is at:

http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-status.pdf

Questions:

Ladislav Vobr - Emirates Telecom Corp - ETISALAT: Is the graph showing
the number of requests per minute?

Axel Pawlik: Yes.

[A bit of a misunderstanding below. That's why questions and answers
don't quite match. Difficult to render - The Scribe].

Carsten Schiefner: Could you clarify whether this is all the e164.arpa
ENUM-related requests or only those to the nameserver(s) under the
RIPE NCC control?

Axel Pawlik: This is only what the RIPE NCC nameserver sees.

Action on RIPE NCC: Statistics from all e164.arpa servers.

Daniel Karrenberg - RIPE NCC: Note that publishing individual zone
statistics would diverge from DNS statistics practice. What Axel shows
is fine. People holding delegations should publish their own stats.

Carsten Schiefner: I meant that we'd like to see secondaries as well.

Peter Koch: It should be a non-issue to get stats from holders of
delegations once ENUM gets used. The worth of tier-0 statistics
numbers is questionable.

Carsten Schiefner: What about the indexation of requests? Everything
used to be available, but now only a list of delegations is published.
The information is still there via deep links, but not linked.

Axel Pawlik: We'll work on it.

Action on Carsten Schiefner: Let NCC know details of missing links.

Carsten Schiefner: Is the service monitored with DNSMON?
Axel Pawlik: Yes.

Carsten Schiefner: The RIPE NCC used to provide IAB with a quarterly
ENUM report, that seems to not be the case any longer. Why?

Action on RIPE NCC: To check whether these reports would still be
necessary and asked for.

E. Jim Reid - DNS-MODA (10 min)
Update on the ETSI ENUM Plugtest


The presentation is at:

http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-etsi.pdf

No question.


F. Peter Koch - DENIC (10 min)
'Walking the 9.4.e164.arpa tree: first results'


The presentation is at:

http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-e164.pdf

Questions:

Peter Koch: Is this useful?

Jim Reid: It's useful to have statistics, but publishing them might
discourage people from investing in ENUM, because the tree is empty at
the moment.

Carsten Strotmann - Men & Mice: Is the bug in all the BIND versions
from 9 to 9.3? You said they aren't used anymore, but it's not the
case.

Peter Koch: I meant that they aren't used in ENUM trials.

Michael Haberler - IPA: Could you make a distinction between numbers
and actually reachable endpoints? And yes, the statistics are useful.

Peter Koch: I kept them in because else the numbers would be even
lower.

Michael Haberler: You could check the reachability with SIP.

Peter Koch: This is interesting. I will contact you.

Carsten Schiefner: Did you apply wildcards on the right and the left
side?

Peter Koch: I stayed DNS-minded, so only on the right side.

Niall O'Reilly: Will you continue with the work and continue to
report?

Peter Koch: Yes, as long as there's something interesting to report.

Action on Peter Koch: Continue and report at RIPE 52.


G. Jim Reid - DNS-MODA (10 min)
E.164 testbed numbers: per country or directly from the ITU?


The presentation is at:

http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-testbed-numbers.pdf

Questions:

Jim Reid: Is reserving a block for a permanent test bed a good idea?

The room: Yes.

Niall O'Reilly: Go to the list.

Peter Koch: A block is probably not sufficient. You should check
RFC2606.

Mark McFadden: It depends what the testbed is for. Individual
countries could have their own blocks.

Patrik Faltstrom - Cisco Systems: It depends what you'll use the
numbers for. To a similar request, the regulator in SE responded that
the telephone system operators managed to do testing without test
numbers.

Jim Reid: It's interesting because then you can get test cases that
you wouldn't get otherwise.

Patrik Faltstrom: But you can request numbers as usual. It depends on
what you're going to test.

Jim Reid: Good point.

Daniel Karrenberg: You need a good set of requirements for what you'll
do. You could go elsewhere than e164.arpa, but implementations might
not be able to deal with that. I agree with Patrik.

Carsten Schiefner: I'm in favour of going through normal channels.

Action on Jim Reid: Stimulate discussion on testbed numbers/ranges.


Y. I/O with other WGs, AOB (10 min)


Carsten Schiefner: From the DNS WG, there's a potential for glitches
in the delegation process. Pre-delegations checks need to be
performed, some results suggest that this is not yet the case.

There's an action on Leo Vegoda in the DNS WG (action item 51.5).
Carsten Schiefner to work out some text and double-check with Leo.

Leo Vegoda: We changed things a few months ago for the reverse tree.
These checks aren't performed automatically for e164.arpa yet, but
manually. It's possible to miss problems. An error occurred, but we're
talking to the holder of the delegation to straighten things.

Niall O'Reilly: How much documentation/input do you need to help you
further?

Leo Vegoda: A document would be fine, but it's probably overkill to go
through the PDP.

Peter Koch: Let's just keep the action on Carsten, keeping this in
mind.

Carsten Schiefner: Is there interest in on-going monitoring of the
quality of the delegations? High quality is important to help
establish alternatives to the traditional methods to transport voice
traffic.

Jim Reid: It depends. On the one hand, the RIPE NCC will support the
integrity and stability of the DNS infrastructure. But the RIPE NCC
isn't supposed to provide monitoring of other services. It could
become an issue of national sovereignity.

Daniel Karrenberg: It's within the reponsibility of the parent to find
out about lame delegations, for instance. This is part of stability.

Peter Koch: I'd like to remind people of issues with '.' (the DNS
root).

AOB: Non-geographic number ranges reachability.

Michael Haberler: We've seen issues with non-geographic numbers. It
takes very long to get these ranges globally reachable. It can be
years. This has happened with Liechtenstein.

Telcos are filtering numbers to make sure that only cheaply reachable
numbers are let through. This is very difficult to debug.

I'll ask the list to test a number from different types of line,
mobile, fixed, etc. Thanks in advance.

Action on all: Respond to AT reachability.


Z. Close (5 min)
. Summarize action items
. Thank scribe and participants


[Meeting closed.]