Your IP Address is: 38.107.191.82
RE: URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary
From: Bound, Jim (Jim.Bound@hp.com)
Date: 09/18/06
- Next message: Bound, Jim: "RE: URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary"
- Previous message: Bound, Jim: "RE: URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary"
- Maybe in reply to: Bound, Jim: "URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary"
- Next in thread: Bound, Jim: "RE: URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary"
John,
Thats funny :----). THis is all pretty simple and here is the rule. To note and ask does this vendors implementation support A, B, or C is fine. THe vendor will jump in and ask the question. When the discussion gets to does this vendor meet the architecture opinions of the NAv6TF is where the line gets crossed even if it is friendly. We have to limit our discussion of any vendor implementation or we run two risks. The most important one is anti-trust values and that form of issue. THe second is any depth or views really to any vendor that supports NAv6TF like Microsoft, Cisco, Juniper , SIs, Providers or others should not be a theme of a mail thread continuously on the list. The risk is we cross a line either in favor or against said vendor. Thus I sent this mail for both IPv6 Forum base tenet reasons and for legal reasons for the TF.
Your second point on the topic of technology I stated clearly should be continued for deployment being DNS, DHCP, Management, Transition etc and that conversation should continue and good actually result in the NAv6TF forming a deployment view from the mail which we can then take a position upon as a body,
Thus if you could start a new thread on that topic and continue the technology analysis of IPv6 analysis that would be great but keep what vendors do what out of the loop.
I hope this helps you and all with the issue.
Thanks
/jim
From: John L Lee [mailto:john@internetassociatesllc.com] Sent: Monday, September 18, 2006 3:28 PM To: Bound, Jim Cc: Bill Cerveny; Christopher L. Morrow; McGuerty, Jay S.; ChrisMitchell; Durand, Alain; Dave Green; Joya Subudhi; nav6tf@ipv6forum.com; moonv6@iol.unh.edu
Subject: Re: URGENT NAv6TF Chair Request : RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary
The Boss has given us orders so -
If there is continued interest in this thread, is there someone who can collect e-mail addresses and then this can continue off list. It would also appear that in the future when issues of interest come up that should not continue on this list that we have a known method to take them off list.
My interest is in the protocol details between client and server implementation in both mixed v4/v6 networks and then v6 only for DNS and DHCP. The behavior issues are also useful.
John Lee
Bound, Jim wrote:
All Members,
I want the rhetoric regarding Vista to stop NOW and refences to any
testing of any specific vendor too. All vendors are affected by this
technology discussion. I encourage continuing this discussion but I
must now ask as Chair to stop now referencing Microsoft Vista it is
totally of out line at this point. It was fine initially and Chris did a
nice thing responding. I could do this with every vendors
implementation on this list, and it is unfair for such NAv6TF discussion
for Vista. Please keep the discussion going but reference technology
and architecture and needs of IPv6 and get off of identifying vendors.
I assume my mail here is crystal clear? If you don't get it send me
private mail but I am traveling so cannot respond as usual but I will.
We do NOT DISCUSS VENDORS or any specific IMPLEMENTATION in the NAv6TF
other a minimal discussion for support and understanding the market
implementation of IPv6. Unless of course the implementation is public
domain code base.
Thank You for your support and understanding, /jim Chair NAv6TF www.nav6tf.org -----Original Message----- From: owner-nav6tf@ipv6forum.comassociated with Teredo.
[mailto:owner-nav6tf@ipv6forum.com] On Behalf Of
Bill Cerveny Sent: Monday, September 18, 2006 12:21 AM To: Christopher L. Morrow; McGuerty, Jay S. Cc: Chris Mitchell; Durand, Alain; Dave Green; Joya Subudhi; nav6tf@ipv6forum.com; moonv6@iol.unh.edu Subject: RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - we ekly summary I'm personally less concerned about the additional DNS request as I am about what assumptions Vista is making about its connectivity (although it is a good thing to clarify that DNS servers won't melt down ;-) ). In my opinion, Vista shouldn't be asking for AAAA records when it has no expectation it will reach the host via IPv6 or ask for an A record when it has no expectation it will reach the host via IPv4. From what Chris Mitchell at Microsoft has said and Dave Green has verified (thanks Chris and Dave!), this appears to be the case, except perhaps for some behavior
To take this a step further, what happens when a Vista box is
connected to an isolated IPv6 network, although it is
connected to the global IPv4 Internet, and can retrieve AAAA
records on DNS for hosts beyond it's isolated IPv6 network
(but reachable via IPv4)? For example if I entered
"www.kame.net" in IE or Firefox from the isolated IPv6
network with global IPv4 connectivity and got a AAAA record
for www.kame.net. With Windows XP, I could experience a
noticeable period where the operating system is figuring out
that it can't reach www.kame.net via IPv6 and finally decides
to reach it via IPv4.
This scenario will certainly happen as large enterprise IPv6
roll-outs occur and for whatever reason, there are intended
or unintended pockets of isolated IPv6 connectivity within
the enterprise. Whenever this behavior occurred with Windows
XP, the rational administrator response was to disable IPv6
on the computer (and the problem did go away). For
IPv6 to be successful on Vista, the operating system must
recover quickly enough that the user doesn't notice any
discernible delay. The user or network administrator
shouldn't assume that a fix to a network problem is to "turn
off" IPv6 on Vista.
Bill
On Mon, 18 Sep 2006 02:05:16 +0000 (GMT), "Christopher L. Morrow"
<christopher.morrow@verizonbusiness.com> <mailto:christopher.morrow@verizonbusiness.com> said:
On Sun, 17 Sep 2006, McGuerty, Jay S. wrote:
moonv6 post from "McGuerty, Jay S." <JAY.S.MCGUERTY@saic.com>
<mailto:JAY.S.MCGUERTY@saic.com> Folks,
In our testing of VISTA we do see
continual period
attempts by VISTA to
find the default Teredo server (and
others). As of
VISTA build 5436 we
simply shut down the Teredo interface and that stops it. We do see the AAAA record I assume your test host wasn't in a domain? (default behaviour was to NOT do terado if there was a domain
association)
lookup first (I personally think it is a great idea) like many of the "sky is falling IPv6 alarms" it seems to me this much to do about nothing. We super sized our DNS server knowing that more of our dual stack environment was going to depend on DNS resolution anyway. you super sizing your DNS server means something completely different to Alain or me... :( There is a concern, and a valid one I believe, though one we'll have to get used itseems.
v/r
Jay
From: owner-moonv6@io.iol.unh.edu on behalf of Chris Mitchell
Sent: Sat 9/16/2006 10:58 AM To: Durand, Alain; green@commandinformation.com; bill@wjcerveny.com; joya@subudhi.com; nav6tf@ipv6forum.com; moonv6@iol.unh.edu Subject: RE: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - weekly summary moonv6 post from Chris Mitchell <cmitch@windows.microsoft.com> <mailto:cmitch@windows.microsoft.com> Alain, The current design in Vista is that Teredo addresses will not be registered in DNS and will not cause AAAA record look ups unless an application is actively using IPv6 via the Teredo tunnel. In other words, "out-of-the-box" Vista systems that are not actively using Teredo will NOT cause AAAA record lookups.
Thanks, Chris Mitchell Group Program Manager -----Original Message----- From: Durand, Alain [mailto:Alain_Durand@cable.comcast.com <mailto:Alain_Durand@cable.comcast.com> <mailto:Alain_Durand@cable.comcast.com> ] Sent: Friday, September 15, 2006 3:00 AM To: Chris Mitchell; green@commandinformation.com; bill@wjcerveny.com; joya@subudhi.com; nav6tf@ipv6forum.com; moonv6@iol.unh.edu Subject: Re: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - weekly summary Could someone clarify how this reconcile with teredo automatic IPv6 connectivity establishment? - Alain. ----- Original Message ----- From: owner-moonv6@io.iol.unh.edu <owner-moonv6@io.iol.unh.edu> <mailto:owner-moonv6@io.iol.unh.edu> To: Dave Green
<green@commandinformation.com> <mailto:green@commandinformation.com> ; Bill Cerveny
<bill@wjcerveny.com>
<mailto:bill@wjcerveny.com> ; Joya Subudhi <joya@subudhi.com>
<mailto:joya@subudhi.com> ;
nav6tf@ipv6forum.com
<nav6tf@ipv6forum.com> <mailto:nav6tf@ipv6forum.com> ;
moonv6@iol.unh.edu
<moonv6@iol.unh.edu>
<mailto:moonv6@iol.unh.edu>
Sent: Thu Sep 14 18:55:43 2006 Subject: [moonv6] RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - weekly summary moonv6 post from Chris Mitchell <cmitch@windows.microsoft.com> <mailto:cmitch@windows.microsoft.com> Thanks for testing and confirming this Dave. This is exactly what we specified as the expected behavior. If anyone has feedback that this isn't right please let us know. Thanks, Chris Mitchell Group Program Manager Windows Networking and Devices -----Original Message----- From: owner-nav6tf@ipv6forum.com
[mailto:owner-nav6tf@ipv6forum.com
<mailto:owner-nav6tf@ipv6forum.com> <mailto:owner-nav6tf@ipv6forum.com> ] On Behalf Of Dave Green Sent: Thursday, September 14, 2006 2:44 PM To: Bill Cerveny; Joya Subudhi; nav6tf@ipv6forum.com; moonv6@iol.unh.edu Subject: RE: Vista DNS behavior - Re: [nav6tf] IPv6 news - weekly summary Bill, et all: We've been aware of this behavior on XP, but had not checked for it in Vista. Since we run a fully dual-stacked environment this behavior wasn't causing us a problem. However, if it was still present it could cause long lags on IPv4-only environments so I tested this today on the Vista Pre-RC1 Build 5536 we are pilot testing for production. To test this, I used an IPv6-capable packet sniffer to capture and decode all DNS queries with Vista on our main dual stacked environment, and on the v4-only and IPv6-only environments we run for application
testing. This is what I found:
-On a native IPv6-dual-stacked
connection Vista Generated 2 DNS
requests for each record and I was able to reach our
IPv6-only content sites.
-In our IPv4-only test environment, I only saw requests for A
records and requests to our IPv6 only content sites failed.
-In our IPv6-only environment, I only saw requests for
AAAA records
and our IPv6 content sites worked fine.
I believe that this behavior is exactly what we wanted
Vista to do.
David Green | VP of R&D | Command
Information
13655 Dulles Technology Drive, Suite 100 | Herndon, VA 20171
O: 703.561.5937 | M: 703.899.9663 |
green@commandinformation.com
-----Original Message----- From: owner-nav6tf@ipv6forum.comVista beta?
[mailto:owner-nav6tf@ipv6forum.com
<mailto:owner-nav6tf@ipv6forum.com> <mailto:owner-nav6tf@ipv6forum.com> ] On Behalf Of Bill Cerveny Sent: Monday, September 11, 2006 6:53 PM To: Joya Subudhi; nav6tf@ipv6forum.com; moonv6@iol.unh.edu Subject: Vista DNS behavior - Re: [nav6tf] IPv6 news - weekly summary Does anyone have any experience with
The two articles below about Vista and DNS melting the Internet
suggest that Vista will always make two DNS requests; one
for the A
record and the other for the AAAA
record. While I might
expect this
to occur if Vista thinks it is connected to an operational IPv6
network, will Vista still do this if it doesn't have a globally
routable IPv6 address or any other
indication that it is
connected to an IPv6-enabled network?
Bill Cerveny
On Mon, 11 Sep 2006 10:38:21 -0400,
"Joya Subudhi"
<joya@subudhi.com>
<mailto:joya@subudhi.com>
said: IPv6 news - weekly summary Main News: Plan for cyberspace available in nearfuture
http://www.ipv6tf.org/news/newsroom.php?id=2237
<http://www.ipv6tf.org/news/newsroom.php?id=2237> <http://www.ipv6tf.org/news/newsroom.php?id=2237>
Other News: IPv6: The future is now http://www.ipv6tf.org/news/newsroom.php?id=2236
<http://www.ipv6tf.org/news/newsroom.php?id=2236> <http://www.ipv6tf.org/news/newsroom.php?id=2236>
Security:
Corporate and Products: Vista's role in Imminent Net Death
overstated
http://www.ipv6tf.org/news/newsroom.php?id=2240
<http://www.ipv6tf.org/news/newsroom.php?id=2240> <http://www.ipv6tf.org/news/newsroom.php?id=2240>
Command Information Appoints Greg Sloan to Lead
Commercial Sector
Business Development
http://www.ipv6tf.org/news/newsroom.php?id=2239
<http://www.ipv6tf.org/news/newsroom.php?id=2239> <http://www.ipv6tf.org/news/newsroom.php?id=2239>
Will Vista stall Net traffic?
http://www.ipv6tf.org/news/newsroom.php?id=2238
<http://www.ipv6tf.org/news/newsroom.php?id=2238> <http://www.ipv6tf.org/news/newsroom.php?id=2238>
Using the Network Diagnostic Tester
http://www.ipv6tf.org/news/newsroom.php?id=2235
<http://www.ipv6tf.org/news/newsroom.php?id=2235> <http://www.ipv6tf.org/news/newsroom.php?id=2235>
This archive was generated by hypermail 2.1.7 : 12/01/06 EST
