On Wed, 18 Dec 2002, Miquel van Smoorenburg wrote:
> Okay. I found /only/ this:
>
> 25-Nov-2002 12:28:44 25-Nov-2002 12:28:44 25-Nov-2002 12:43:44
> R:cwu4@??? 0 77 SMTP error from remote mailer after RCPT TO:<cwu4@???>: host uni00mr.unity.ncsu.edu [152.1.
>
> 25 november ?
That's so old, it should have been ignored. (Clearly it hasn't been
updated recently.)
> after about an hour the retry database shows:
>
> 18-Dec-2002 14:19:24 18-Dec-2002 14:19:24 18-Dec-2002 14:34:24
> R:cwu4@??? 0 77 SMTP error from remote mailer after RCPT TO:<cwu4@???>: host uni03mr.unity.ncsu.edu [152.1.
>
> .. and sure enough, in the mean time all messages that were stuck
> have now been bounced back to the sender.
So the problem is a temporary error just on that one address. That does
give rise to routing delays. However, there should be a timeout in the
end.
> So I guess basically this is a 'retry database corrupt' issue ?
No, I don't think so. It doesn't look corrupt. (The error message is
truncated at a fixed length to save space; it's just for information.)
> We did
> have one or two powerdips in Amsterdam in the last few weeks .. so
> that would be a possibility. As I understand from the docs, it is
> harmless to remove the retry database, so perhaps you should tell
> people like me right away to nuke the retry database....
Yes, it is always harmless to remove the database. But there are subtle
problems from time to time, which is why I like to understand what's
going on. In fact, this looks rather like a bug that I have just fixed
in 4.12. (However, ISTR you are still on Exim 3, and I don't think the
circumstances are the same there.)
Philip
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
