Re: [Exim] Message frozen but not frozen ?

トップ ページ
このメッセージを削除
このメッセージに返信
著者: Philip Hazel
日付:  
To: Ian Jackson
CC: exim-users
題目: Re: [Exim] Message frozen but not frozen ?
On Wed, 8 Sep 1999, Ian Jackson wrote:

> I disagree - it's usually a good idea to write log records
> etc. first. Otherwise there's a risk that a message would be frozen
> but noone would be told, which would be bad.


Come to think of it, maybe that's why it is the way it is. I can't
remember - this is one of the oldest bits of Exim code. However, even
keeping the order, there is still room for improvement because at
present "Frozen" is logged when the address that caused it is processed
- but there may be other addresses to be processed before the header
file is actually updated.

> Possibly you want to suspend certain signals (HUP and TERM are
> traditional) at certain points. A really paranoid thing might be
> always to disable TERM inside things that should be `atomic'.


HUP is, I think, completely suspended except for the daemon process.
I'll have a look at whether suspending TERM might be helpful. Thanks.

-- 
Philip Hazel            University of Cambridge Computing Service,
ph10@???      Cambridge, England. Phone: +44 1223 334714.