OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: Alexandre Dulaunoy (alex_at_conostix.com)
Date: Tue Oct 01 2002 - 01:44:38 CDT

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    dbmail seems really a good idea to solve file system issue and common
    backend issue.

    Do you have a big installation running dbmail with Postfix in front ?

    Are you still running the smtp from dbmail ? or only Postfix ? Is
    there a patch for Postfix in that case ?

    adulau

    On Mon, 30 Sep 2002, dingomicrobsd.net wrote:

    > While the url is useful to a degree the most simple way to build a
    > cluster of MRAs is to use postfix and dbmail with mysql4. these three
    > combined with anti-virus give you a true scalable redundent mail system
    > for both pop/imap and the key to it all being the mysql database and
    > replication. dbmail allows us to manage downtime and system failures in
    > a more timely manner. and it is known to be scalable to tens of
    > thousands of users, uses no accounts on the system, centralizes mail
    > storage in the database and requires no /var/mail to hold users mail as
    > the messages themselves are stored in the database.
    >
    > Clifton Royston wrote:
    >
    > >On Mon, Sep 30, 2002 at 02:48:08PM +0200, Marco Corte wrote:
    > >
    > >
    > >>Domingo Antonio wrote:
    > >>
    > >>
    > >>>helo all!
    > >>>where can i find about postfix clusters?
    > >>>i trying to create a cluster of postfixs..
    > >>>
    > >>>
    > >>I found http://uni.x-si.org/hamail.txt very useful.
    > >>
    > >>
    > >
    > >This "punts" (i.e. says "we'll solve this later") on the most difficult
    > >part of the problem, and in some sense the only difficult part of the
    > >problem: namely sharing the user mail store which is dynamically
    > >updated by POP/IMAP sessions.
    > >
    > >Creating a cluster of MTAs is easy. As someone else noted, that's what
    > >MX records are for. A cluster of MRAs (mail retrieval agents) is the
    > >hard problem.
    > >
    > >For this you need a working high-reliability NFS server, and you need a
    > >reliable NFS client implementation in the OS on your mail servers.
    > >(Note please that I am not talking about using NFS for queues, only for
    > >the user mail store, be it maildir or what have you.)
    > >
    > >
    > >
    > >>>when 1 is down another do the job.. ( and all data in down server is copied to
    > >>>the another.. )
    > >>>
    > >>>
    > >>How can you do it? If a server crashed, how do you access the files on
    > >>its disks? You need to share resources between hosts which is an OS issue...
    > >>
    > >>
    > >
    > >Yes, and you need to maintain the coherence of those resources while
    > >they are being updated.
    > > -- Clifton
    > >
    > >
    > >
    >
    >
    >
    > -
    > To unsubscribe, send mail to majordomopostfix.org with content
    > (not subject): unsubscribe postfix-users
    >
    >
    >

    -- 
    			      Alexandre Dulaunoy -- http://www.foo.be/
      3B12 DCC2 82FA 2931 2F5B 709A 09E2 CD49 44E6 CBCD  ---   AD993-6BONE
    "People who fight may lose.People who do not fight have already lost."
    							Bertolt Brecht
    

    - To unsubscribe, send mail to majordomopostfix.org with content (not subject): unsubscribe postfix-users