Re: [EAI] Downgrade testing

----- Original Message ----- 
From: "Ernie Dainow" <edainow@ca.afilias.info>
To: "YAO Jiankang" <yaojk@cnnic.cn>
Cc: "Stephane Bortzmeyer" <bortzmeyer@nic.fr>; "EAI" <ima@ietf.org>; <abelyang@twnic.net.tw>; <meldin78@hotmail.com>
Sent: Friday, August 07, 2009 7:52 PM
Subject: Re: [EAI] Downgrade testing - Results

> Some of your failures may be due to the environment and not the 
> Downgrade. I ran a 'base' test by sending the email from an ASCII 
> address to try and identify which failures are due to the environment 
> (firewalls, network, etc.). 
> 

today,  we use the ascii address from the same machine to these 10 addresses. all get success.

so, that is to say that it is not the environment problem. 

Yao Jiankang
CNNIC

> On the Downgrade test, I received replies from all servers except these
> echo@nic.fr
> echo@tu-berlin.de
> echo@cnam.fr (this address does not exist and you get a reject from  
> echo@copernic.cnam.fr)
> 
> You received replies from the first two but not from three others that I 
> did receive a reply from. This is a surprising difference. A base test 
> might help explain it.
> 
>    -Ernie Dainow
> 
> 
> YAO Jiankang wrote:
>> btw, the implmentation is based on postfix.
>> could TWNIC JPRS NIDA do the similar tests and report the results?
>>
>> YAO Jiankang
>> CNNIC
>>
>> ----- Original Message ----- 
>> From: "YAO Jiankang" <yaojk@cnnic.cn>
>> To: "Ernie Dainow" <edainow@ca.afilias.info>; "Stephane Bortzmeyer" <bortzmeyer@nic.fr>; "EAI" <ima@ietf.org>
>> Sent: Thursday, August 06, 2009 11:10 AM
>> Subject: Re: [EAI] Downgrade testing - Results
>>
>>
>>   
>>> we send to the following 10 address, with <eai<ascii>>
>>>     
>>>>> * echo@generic-nic.net 
>>>>> * echo@nic.fr 
>>>>> * Echo@TU-Berlin.DE 
>>>>> * echo@tu-chemnitz.de 
>>>>> * echo@ouain.com 
>>>>> * repondsmoi@crdp.ac-versailles.fr
>>>>> * echo@cnam.fr
>>>>> * ping@stamper.itconsult.co.uk 
>>>>> * ping@oleane.net
>>>>> * check-auth@verifier.port25.com
>>>>>         
>>> we get the nice response from the following 6 address
>>> echo@generic-nic.net
>>> echo@nic.fr
>>> Echo@TU-Berlin.DE
>>> echo@ouain.com
>>> check-auth@verifier.port25.com
>>> repondsmoi@crdp.ac-versailles.fr
>>>
>>>
>>> we get the reject information from 1 address
>>> echo@copernic.cnam.fr
>>>
>>>
>>> we can not get any response from other 3 addresses.
>>>
>>>
>>>
>>> Yao Jiankang
>>> CNNIC
>>>
>>>
>>>
>>>
>>>
>>> ----- Original Message ----- 
>>> From: "Ernie Dainow" <edainow@ca.afilias.info>
>>> To: "Stephane Bortzmeyer" <bortzmeyer@nic.fr>; "EAI" <ima@ietf.org>
>>> Sent: Thursday, July 09, 2009 9:09 PM
>>> Subject: Re: [EAI] Downgrade testing - Results
>>>
>>>
>>>     
>>>> These echo tests were a good suggestion. I ran a base test from an ASCII 
>>>> address against all echo servers in the list. All responded except 
>>>> echo@cnam.fr. Email bounced with "unknown address", so I  eliminated 
>>>> that server from the test, leaving a total of 9 echo servers.
>>>>
>>>> The success rate for receiving a response from echo servers:
>>>> 1) From ASCII, no downgrade      9/9      100%
>>>> 2) From EAI, downgrade               7/9        78%
>>>>
>>>> auth-results@verifier.port25.com provided some SpamAssassin analysis:
>>>> 1) BODY: Bayesian spam probability is 1 to 5%
>>>> 2) BODY: Bayesian spam probability is 5 to 20%
>>>>
>>>> For the downgrade test sent to individual participants the results were:
>>>> Total Sent    11   (all on different email domains)
>>>> Received       9    82%
>>>> Blocked        2    18%
>>>>
>>>> These are not large samples, but it is interesting to see that the 
>>>> success rate in both tests were similar, and were in the SpamAssassin 
>>>> range of 20% probability of spam.
>>>>
>>>> I encourage other implementers to run the same set of tests so we can 
>>>> get a larger sample and see if there are differences between Downgrade 
>>>> implementations.
>>>>
>>>>    -Ernie
>>>>
>>>>
>>>> Stephane Bortzmeyer wrote:
>>>>       
>>>>> On Mon, Jun 22, 2009 at 11:19:48AM -0400,
>>>>>  Ernie Dainow <edainow@ca.afilias.info> wrote 
>>>>>  a message of 41 lines which said:
>>>>>
>>>>>   
>>>>>         
>>>>>> The Downgrade test reported on the wiki basically tested against gmail.  
>>>>>> It would be valuable to do more testing 'in the wild' to see how  
>>>>>> Downgrade traverses various email systems in other organizations.
>>>>>>     
>>>>>>           
>>>>> You can test against various auto-responders which send you back your
>>>>> message (if you know other email addresses of auto-responders, do not
>>>>> hesitate to publish them).
>>>>>
>>>>> * echo@generic-nic.net 
>>>>> * echo@nic.fr 
>>>>> * Echo@TU-Berlin.DE 
>>>>> * echo@tu-chemnitz.de 
>>>>> * echo@ouain.com 
>>>>> * repondsmoi@crdp.ac-versailles.fr
>>>>> * echo@cnam.fr
>>>>> * ping@stamper.itconsult.co.uk 
>>>>> * ping@oleane.net
>>>>> * check-auth@verifier.port25.com
>>>>>   
>>>>>         
>>>> _______________________________________________
>>>> IMA mailing list
>>>> IMA@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/ima
>>>>       
>>> _______________________________________________
>>> IMA mailing list
>>> IMA@ietf.org
>>> https://www.ietf.org/mailman/listinfo/ima