Re: IPv6 ready Logo program Self test (HOST only) version 1.3.6

Dear Mr. Bellino and IPv6 folks,

I am also working on passing Tahi, but for Phase 2. And I've had very 
helpful replies from the Tahi group on issues about their tests and Perl 
code.

What the rules of the tests are, will become an issue for us, so I am 
interested in your thread.

My working interpretation of the rules is that all the data and timing 
in the Perl code within a test are "official Tahi rules." I have 
questioned the Tahi group about particular tests and gotten helpful and 
reasoned replies, and still have outstanding questions about some tests.

It is reasonable to ask, as you have done, "does this test pass correct 
results?" In one instance the Tahi reply was, "we will fix this problem 
in a later version."
You could try sending this particular issue (4) to [email protected].

I have altered Tahi Perl code that affects the system between tests. My 
assumption is that I can send in my ct/ directories with results that 
have been run with pure Tahi and still pass, even if the delays between 
tests were under my control and I communicated from the TN to the NUT to 
signal neighbor cache purges. The timing of neighbor cache purges is 
left open by the RFCs, but critical for passing the tests.

Tx,
Tim McFadden
--------------------------------------
Interniche Technologies, Inc.
51 East Campbell Avenue, Suite 160
Campbell, CA 95008
408-540-1160x207
[email protected]
www.iniche.com
--------------------------------------


Bellino, Phil wrote:

>Hello,
>I am executing the IPv6 Ready Logo Program Self test version 1.3.6.  I have
>accomplished the following:
>1.  The Ipv6 Specification tests all passed.
>2.  The ICMPv6 tests all passed.
>3.  The Neighbor Discovery tests all passed.
>4. I have an issue with the Stateless-autoconf tests. There are 26 tests and
>tests 1-23 all passed.  Test 24-26 hang because one or the other "VSLEEP"
>formulas end out with a negative value for seconds to wait(which causes the
>hangs).
>
>My questions are:
>1.  Has anyone seen this in testing?
>
>2.  Are we allowed to change the pearl scripts with respect to the vsleep
>formulas to make them work as the comments read in the respective .seq
>files? RA_sameVLT.seq for test 24, RA_incVLT.seq for test 25, and
>RA_decVLT.seq for test 26.  
>
>3.  Actually, I have done this, but isn't this also breaking the rules of
>the tests?
>
>Thanks,
>Phil Bellino
>
>============================
>Phil Bellino
>MRV Communications, Inc.
>Boston Product Division
>295 Foster St.
>Littleton,MA 01460
>Tel: (978)952-4807
>Email: [email protected]
>============================
>
>  
>

---------------------------------------------------------------------
The IPv6 Users Mailing List
Unsubscribe by sending "unsubscribe users" to [email protected]