QUESTION: About the RFCs of DCCP

QUESTION: About the RFCs of DCCP

[Wei Yongjun]

Hello,

I am plan to make an Conformance Test Suite for DCCP which is under
packet level. It's aim to test the packet format, options, feature,
sequence numbers, event processing etc. base on the rfc4340. Maybe also
rfc4341, rfc4342.
It is there some fix RFCs or fix draft for the rfc4340? Or base on
rfc4340 is enough?

Wei Yongjun

Re: QUESTION: About the RFCs of DCCP

[Ian McDonald]

On Fri, May 23, 2008 at 4:55 PM, Wei Yongjun <yjwei@cn.fujitsu.com> wrote:
> Hello,
>
> I am plan to make an Conformance Test Suite for DCCP which is under
> packet level. It's aim to test the packet format, options, feature,
> sequence numbers, event processing etc. base on the rfc4340. Maybe also
> rfc4341, rfc4342.
> It is there some fix RFCs or fix draft for the rfc4340? Or base on
> rfc4340 is enough?
>
> Wei Yongjun
>
The better place to ask this is on the IETF mailing list. See here for
more details:
http://www.ietf.org/html.charters/dccp-charter.html

-- 
Web: http://wand.net.nz/~iam4/
Blog: http://iansblog.jandi.co.nz

Re: QUESTION: About the RFCs of DCCP

[Gerrit Renker]

| I am plan to make an Conformance Test Suite for DCCP which is under
| packet level. It's aim to test the packet format, options, feature,
| sequence numbers, event processing etc. base on the rfc4340. Maybe also
| rfc4341, rfc4342.
| 
Maybe this is of help, Andrea Bittau did something similar earlier, at
http://tapir.cs.ucl.ac.uk/pccd/

Just out of curiosity - are you using an existing system such as PCS 
(http://pcs.sourceforge.net/) or KAME/TAHI?

Re: QUESTION: About the RFCs of DCCP

[Wei Yongjun]

Gerrit Renker wrote:
> | I am plan to make an Conformance Test Suite for DCCP which is under
> | packet level. It's aim to test the packet format, options, feature,
> | sequence numbers, event processing etc. base on the rfc4340. Maybe also
> | rfc4341, rfc4342.
> | 
> Maybe this is of help, Andrea Bittau did something similar earlier, at
> http://tapir.cs.ucl.ac.uk/pccd/
>   
I have saw this, but it is has so little test case. And looks like it 
has no test log detail.
> Just out of curiosity - are you using an existing system such as PCS 
> (http://pcs.sourceforge.net/) or KAME/TAHI?
>
>   
I am used TAHI's v6eval, and added DCCP support. So The test log  looks 
very good.^_^

Now it is under development. After it completed, I'll publish the test 
tools and test result.

Re: QUESTION: About the RFCs of DCCP

[Gerrit Renker]

> I am used TAHI's v6eval, and added DCCP support. So The test log  looks  
> very good.^_^
>
> Now it is under development. After it completed, I'll publish the test  
> tools and test result.
>
That sounds great and would be a very useful tool to have. Can you
please drop us a line if something is available?

- Gerrit

Re: QUESTION: About the RFCs of DCCP

[Wei Yongjun]

Gerrit Renker wrote:
>> I am used TAHI's v6eval, and added DCCP support. So The test log  looks  
>> very good.^_^
>>
>> Now it is under development. After it completed, I'll publish the test  
>> tools and test result.
>>
>>     
> That sounds great and would be a very useful tool to have. Can you
> please drop us a line if something is available?
>   

It's a pleasure to this.

At this time, the TAHI support is completed for rfc 4340, and I am 
making the test frame for DCCP test, and it has little test case in it 
although the test frame is under test.

It also found something does not completed implement rfc, such as the 
two patch I had send to the mail list, and the BUG [Re: Reporting 
possible bug in DCCP Feature negotiation]  which is reported by Leandro 
Sales can be reproduct simplely.

Re: QUESTION: About the RFCs of DCCP

[Gerrit Renker]

> It also found something does not completed implement rfc, such as the  
> two patch I had send to the mail list, and the BUG [Re: Reporting  
> possible bug in DCCP Feature negotiation]  which is reported by Leandro  
> Sales can be reproduct simplely.
>
Then it would be even more valuable to have such a tool. The bug you
referred to was hard to track down manually since it only occurred when
the Ack carrying the Confirm options was lost. I could only reproduce it
by modifying the kernel to drop the single Ack of the client completing
the 3-way handshake. With a tool, this would have been much easier.

Please do report any faults, omissions or incompleteness that you find
while testing.

- Gerrit