Re: [Exim] quota_warn_threshold specifics

Page principale
Supprimer ce message
Répondre à ce message
Auteur: Mark T. Valites
Date:  
À: Christian Balzer
CC: exim-users
Sujet: Re: [Exim] quota_warn_threshold specifics
On Tue, 22 Jun 2004, Christian Balzer wrote:

> I think you want to have a wishlist item where you can feed external
> quota information from a helper script/program into exim. You might be
> able to hack something up with quota_size_regex and manipulating the
> filenames to match the FS respective quota they are consuming, but that
> is uglier than I want to imagine things. :)


Yes, I think that's exactly what I'd like.

But isn't that essentially what I have by calling the perl subroutine
mentioned in the parent of this thread to return the about of used disk
space?

I also found also found this setting last night:

quota_directory
Type: string, expanded
Default: unset

This option defines the directory to check for quota purposes when
delivering into individual files. The default is the delivery directory,
or, if a file called maildirfolder exists in a maildir directory, the
parent of the delivery directory.

Used in conjuction with the perl, wouldn't this buy me what I want? I know
it's not perfect - are there any places it could go horribly wrong, or
reasons why this would be 'A Really Bad Idea'? (I'm suspecting there
are... overhead and computation time to add up the size of all the files
in a directory come to mind). Does anyone else use the quota_directory
setting or maildirfolder?

If it is a bad idea, would it be so bad to use the percentage of used
blocks returned from the other perl subroutine as a conditional on a
router?

Just looking for thoughts here...

> >-The documentation mentioned that retry rules were able to determine if a
> >termporary delivery failure was the result of a user being over quota. Is
> >this just Exim quotas? or will it be able to detect a file system quota
> >problem?
> >
> Exim quotas, a file system quota incident will just look like a failure
> to write the file out and bounce accordingly.


Oh well...

> >-Secondly, how exactly do to the retry rules determine when a mailbox was
> >last read?
> >
> The retry rules are working on the queue and the hints DBs there, the
> mailbox size is tested each time a delivery opportunity according to
> the retry rules comes up.
> (things are bit more complex with maildir_use_size_file, but for all
> purposes the same principles apply)


Forgive me if I'm being thick, but I'm not sure I follow - maybe I wasn't
clear with my original question. It appears I can tweak my retry rules to
say:

'if a user is over quota and hasn't checked their email in 6 months, set
the retry times to 0'

begin retry
*    quota_180d


How does Exim determine when the last time a user checked their mail? (I
realize I may not be able to utilize this if I can't get past the problems
above, but I'm still curious.)

--
Mark T. Valites
Unix Systems Analyst
Computing & Information Technology
SUNY Geneseo
>--))> >--))>