Problems with Sieve

[John Alton Tamplin]

Peter Burggraef wrote:

>>What version of db are you running?  On Solaris 9, I had similar
>>problems (open returning ENOSPC) after a few weeks of uptime with db
>>4.1.24/25, which went away after downgrading to 4.0.14.
>>    
>>
>How can I find out the version?
>  
>
I don't use Debian, so I can't tell you how to find out the version of 
packages there.

>What can I do, if I don't want to downgrade?
>  
>
Wait for someone to fix the code or fix it yourself -- I haven't dug 
into the db code, but since it breaks after running for a while and an 
older version doesn' t have the same problem, it looks like some change 
between those versions introduced some sort of resource leak.

>>Your duplicate delivery database is corrupted.  If it is unavailable for
>>whatever reason, sieve can't process any messages.  You need to recover
>>your duplicate delivery database, or just delete it and recreate it
>>since it doesn't hold anything that will cause a lot of pain if it is
>>lost (your users may see duplicate copies of messages from mailing lists
>>that are still in flight).
>>    
>>
>I think that happend, when I restarted the service. Is that true?
>  
>
In my case, stopping and restarting usually fixed the problem.  I had 
assumed your logs were from initially starting the service, so I 
expected that meant the files themselves were trashed (which also 
happened to me a couple of times) and needed to be rebuilt or recovered 
from tape.  After looking back at your logs, I see those errors were 
during message delivery, so if restarting clears up your problem than 
you should be fine other than having to periodically restart the server 
to avoid the problem (and of course the loss of filtering until you do).

-- 
John A. Tamplin                               Unix System Administrator
Emory University, School of Public Health     +1 404/727-9931