Call for INN 2.5.2 testing (new NNTP RFC 3977)

Ray Banana rayban at banana.shacknet.nu
Thu Nov 19 05:22:18 UTC 2009


Julien ÉLIE wrote:

>> The closer I look the more I get confused:
>>
>> Nov 17 04:47:55 xxx.xxx.xxx.xx nnrpd[8520]: filter: Checking message 
>> <slrn20091109075439.943.twooffive at may.eternal-september.org>
>> Nov 17 04:47:55 xxx.xxx.xxx.xx innd: localhost:172 bad_messageid 
>> <slrn20091109075439.943.twooffive at may.eternal-september.org>
>> [...]
>> Nov 17 13:59:57 188.40.43.213 nnrpd[3373]: nnn.nnn.nnn.nnn post ok 
>> <slrn20091109075439.943.twooffive at may.eternal-september.org>
>>
>>
>> This explains why the article exists on the server, but not why it was 
>> rejected on the first attempt and accepted on the second.
>> Looking at the time, I assume that at 04:47:55 the server was still 
>> running inn-STABLE-20091115 and I changed to inn-STABLE-20091116
>> some time between 05:00 and 05:30.
> 
> The new parser was already in inn-STABLE-20091115 (but not before).
> inn-STABLE-20091116 only added logs at two more places in the code.
> So it should have changed nothing.
> 
> It seems like Mike had problems with Eternal-September only one time
> (he posted through aioe):
> 
> Date: Sun, 15 Nov 2009 18:00:29 -0500
> Message-ID: <slrn20091115175936.69.twooffive at may.eternal-september.org>

That was a different issue and it happened about 36 hours before I
updated the server to inn-STABLE-20091116.

> Did his last post generate a bad_messageid on your server?

None of his subsequent postings generated a bad_messageid error,
although his message ids always follow the same pattern:
slrn generates the local part and uses the E-S FQDN as the domain part.

As this was a singular phenomenon that occured under undefined 
conditions, I think it does not provide enough information for an error 
analysis.

Anyway, these are the error messages from yesterday's log:

Nov 18 02:30:40 188.40.43.213 innd: news.news4us.nl:91 bad_messageid 
<4b034e3d$1 at news.>
Nov 18 05:37:28 188.40.43.213 innd: news.news4us.nl:212 bad_messageid 
<4b037a00$1 at news.>
Nov 18 12:59:25 188.40.43.213 innd: news.news4us.nl:154 bad_messageid 
<4b03e196 at news.>
Nov 18 19:06:55 188.40.43.213 innd: news.trigofacile.com:372 
bad_messageid <hi-Ray!>

;-)

All message ids are obviously invalid and the error message is correct.

I have excluded ~ 300 error messages of the 
<C724DEE1.4B88E%nicknaym@[remove_this].gmail.com>
type of message id.





More information about the inn-workers mailing list