Re: Migrate 200 Million of rows

From: Holger Baer <holger.baer_at_science-computing.de>
Date: Thu, 31 Jul 2003 09:17:03 +0200
Message-ID: <3F28C26F.8020901@science-computing.de>

Billy Verreynne wrote:
> Norman Dunbar <Norman.Dunbar_at_lfs.co.uk> wrote
>
>

>>What I really like about lossless compressors is the fact that you can
>>compress a file, then compress it again and again and again until you
>>end up with a 1 KB file (roughly). This can be copied to a floppy and
>>your entire database cold backup sent to whoever you like. No need for
>>all that NFS stuff - just a few passes through the compressor program
>>does the trick.

>
>
> Yeah.. we should market these neato tricks. I also have one. It deals
> with backups. I have discovered this amazing device file on Unix. I
> have backed up an entire physical database to it. But the device stays
> the same size. I'm sure it is implementing this compression method
> you're speaking of. I have not yet figured out how to read from the
> device.. but I'm getting there. In the meantime I have made the
> symbolic link /dev/tape0 for this wonderous device called /dev/null...
>
> <GD&RLH>
> --
> Billy

You forgot to mention the amazingly reduced backup time! So even if you can't read back from it, your backup window should drop down to a matter of a few minutes.

IIRC, there was even an article in a german magazine a few years back, where they tested a hardware null device which worked even faster than the usual software solution. Mind you, I don't remember the exact year, but I think it was the april issue ;-)

Cheers

Holger Received on Thu Jul 31 2003 - 02:17:03 CDT