but wait, there is more!. :) in squirrelmail try moving a big email from inbox to some folder. open that folder and delete the email. quota usage won't change as it should. effectively, deletion from any folder other than INBOX will not change quota usage.
there is plenty of inconsistencies like that all over the place. I realized that taking care of quota recalculation and other maintenance activities should be performed via scripts. -- Igor On Fri, Mar 11, 2005 at 04:06:41PM -0500, Charles Sprickman wrote: -> On Fri, 11 Mar 2005 [EMAIL PROTECTED] wrote: -> -> >this is a listing of a sample Maildir directory on my server. it has all -> >these files and directories that have to do with courier IMAP. while -> >they don't take much in some cases, sometimes they can take a couple of -> >percent from the overall quota. in addition there are two directories -> >(IMAP folders) .Junk (where all spam is forwarded) and .Trash (courier -> >IMAP/squirrelmail default trash +folder). -> -> Doh! Now I see why I'm off so much. I forgot about .Trash not being -> counted. I'll have to revamp things to account for that. -> -> Thanks! -> -> Charles -> -> -> >-- -> >Igor -> > -> > -> >On Thu, Mar 10, 2005 at 07:26:10PM -0500, Charles Sprickman wrote: -> >-> On Wed, 9 Mar 2005 [EMAIL PROTECTED] wrote: -> >-> -> >-> >this method is of course fast and dirty, but it might not work for -> >-> >everyone out there. -> >-> > -> >-> >the best aproach is to write a perl script that runs queries on a -> >-> >database, gets path to a user's mailbox on the server, uses File::Find -> >-> >to include only files in Maildir/cur and Maildir/new as well as -> >-> >:subfolders:/cur and :subfolders:/new with the exception of trash -> >-> >folders. that should give a much clearer picture. in addition it can -> >-> >scale better. -> >-> -> >-> I have something similar, but it's not very generic, so I don't know if -> >-> anyone would be interested. I've added a few fields to the vpopmail -> >table -> >-> to record when they first hit quota so I can get a report (quick php -> >page) -> >-> of who's over and for how long. -> >-> -> >-> To get usage, I'm just hitting the user's homedir with du, can you -> >comment -> >-> a bit on why I would need to count only the mail? There shouldn't -> >really -> >-> be anything of note in the user's directory besides the mail... -> >-> -> >-> It is quite interesting to see how much the "real" usage (provided by -> >du) -> >-> can differ from the maildirsize info. -> >-> -> >-> Charles -> >-> -> >-> >-- -> >-> >Igor -> >-> > -> >-> > -> >-> >On Wed, Mar 09, 2005 at 08:08:10AM -0800, Charles J. Boening wrote: -> >-> >-> Here's what I do. I don't have too many domains or users do it -> >doesn't -> >-> >-> take too long. -> >-> >-> -> >-> >-> du -sk /home/vpopmail/domains/*/*/Maildir > /tmp/maildirusage && du -> >-sk -> >-> >-> /home/vpopmail/domains/*/*/*/Maildir >> /tmp/maildirusage && cat -> >-> >-> /tmp/maildirusage | sort -n | tail -n50 | sort -n -r | mail -s "Mail -> >-> >-> User Quota" <your-email-here> -> >-> >-> -> >-> >-> -> >-> >-> This only gets the top 50 users. You can change it easily. I don't -> >-> >-> think it would be very efficient for multiple domains. In that -> >case you -> >-> >-> may want to break it up into a run for each domain. -> >-> >-> -> >-> >-> I'd definitely like to see something more efficient :) -> >-> >-> -> >-> >-> -> >-> >-> Charlie -> >-> >-> -> >-> >-> -> >-> >-> -> >-> >-> ________________________________ -> >-> >-> -> >-> >-> From: Tanmaya Anand [mailto:[EMAIL PROTECTED] -> >-> >-> Sent: Tuesday, March 08, 2005 11:44 PM -> >-> >-> To: vchkpw@inter7.com -> >-> >-> Subject: [vchkpw] listing over-quota users -> >-> >-> -> >-> >-> -> >-> >-> -> >-> >-> Hi -> >-> >-> -> >-> >-> I need to list all over-quota users on the system for all -> >-> >-> domains. -> >-> >-> Can anyone help me with some similiar perl / shell script to -> >get -> >-> >-> this done. -> >-> >-> -> >-> >-> Regards, -> >-> >-> Tanmaya -> >-> >-> -> >-> >-> -> >-> >-> -> >-> >-> <http://clients.rediff.com/signature/track_sig.asp> -> >-> >-> -> >-> >-> -> >-> > -> >
