Re: Archiving of big mailboxes
On Sun, Apr 14, 2002 at 02:16:31PM -0700, Brian Nelson wrote:
> Grzegorz Prokopski <greg@sente.pl> writes:
>
> > Hello!
> >
> > I thought some of You can have a problem which is similar to mine.
> > I am subscribed to many, many lists and I get more than 200 emails
> > every day (maybe more, I didn't check exactly).
> > I use procmail to automaticly sort this mail to different files.
> > Then I use IMAP server and various clients (NN, Evolution, mutt) to
> > acces my mail.
> >
> > The problem is that the mailboxes grow and grow steadily and it takes
> > more time to check for new mail in every of them, more time to get
> > message indexes etc. And - to be honest - I'd like to backup those
> > 100MB+ files to some CD or at least compress them. But You need
> > to cut out the "older" part from them first.
...
> gnus, total expiry. It's all automatic. I just delete the stuff since
> all my mailing lists have web archives, but you can make it expire to
> another folder.
>
> Back when I used mutt, I used
>
> folder-hook "+lists" 'push "<delete-pattern>~d >2w<enter>"'
>
> which marks everything in folders matching "+lists" older than 2 weeks
> for deletion.
Sounds like he wants to keep the old mail, which is what I do.
I use mutt to read the mail, and (depending on volume) once
a month, once a quarter, once every half-year or once a year
I select all messages older than a (month, quarter, half-year,
or year) and save them all to another folder. Then you can copy
that folder off to a cd if you like, and/or compress it.
I also use procmail to pre-sort the mail into topics.
So: mail comes in. procmail pre-sorts it into topics.
Then when I read the various folders, once in a while I will
save all the older messages to a separate folder with the date
and time-period in the name: eg: debian-user-200203 for
the March 2002 debian-user messages. My mutt configuration
automatically marks those messages in the debian-user mail
folder as deleted, and I can delete them safely because they
now exist in the other folder (debian-user-200203). I usually
go to a shell and compress the older folders:
$ gzip debian-user-200203
$ ls debian-user*
debian-user debian-user-200010.gz debian-user-200108.gz
debian-user-200001.gz debian-user-200011.gz debian-user-200109.gz
debian-user-200002.gz debian-user-200012.gz debian-user-200110.gz
debian-user-200003.gz debian-user-200101.gz debian-user-200111.gz
debian-user-200004.gz debian-user-200102.gz debian-user-200112.gz
debian-user-200005.gz debian-user-200103.gz debian-user-200201.gz
debian-user-200006.gz debian-user-200104.gz debian-user-200202
debian-user-200007.gz debian-user-200105.gz debian-users
debian-user-200008.gz debian-user-200106.gz
debian-user-200009.gz debian-user-200107.gz
Oops I see I've saved one or a few messages to the wrong
list name (debian-users). I'll have to clean that up some day.
Once they are compressed you could copy them to a cd and
erase them from your hard disk.
Other lists don't accumulate so quickly and I save them every
three months, every 6 months, or every year.
debian-sparc-2001q1.gz netbsd-users-2000h2.gz sparc-list-2000h1.gz
debian-sparc-2001q2.gz netbsd-users-2001h1.gz sparc-list-2000h2.gz
debian-sparc-2001q3.gz pilot-2001h1.gz sparc-list-2001h1.gz
hurd-2000h1.gz pilot-2001h2.gz sparclinux-2001h1.gz
hurd-2000h2.gz port-sparc-2000h1.gz sparclinux-2001h2.gz
hurd-2001h1.gz port-sparc-2000h2.gz suns-at-home-2000h1.gz
netbsd-help-2000h1.gz port-sparc-2001h1.gz suns-at-home-2000h2.gz
netbsd-help-2000h2.gz port-sparc64-2001h1.gz suns-at-home-2001h1.gz
netbsd-help-2001h1.gz spam-2000h1.gz tech-userlevel-2000h1.gz
netbsd-help-2001h2.gz spam-2000h2.gz tech-userlevel-2000h2.gz
netbsd-users-2000h1.gz spam-2001h1.gz tech-userlevel-2001h1.gz
To grep through old messages, you can use:
$ zcat hurd-*gz | grep what_youre_looking_for | less
----- again, at the command line.
I just do this for high-volume lists. When a list gets to be
"too big" (ie takes too long to load into mutt), then I whack off
a piece of history and compress it with the above procedure.
I do this because I have a slow dial-up connection, and this is
one way to be able to look stuff up before asking a question on the
mailing list. Also, who knows when the net will cease to be free
(in either the beer or speech meaning)?
--
bjb@achilles.net
Welcome to the GNU age! http://www.gnu.org
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: