>Hans-Ole Kaae writes:
>
>>(I'm thinking on how much you can expect to hold on one tape - using both
>>the built-in hardware compression
>>plus ORBiTs software ditto).
>
>Using both hardware and software compression will result in a decrease in
>performance, capacity or both. The reason is that compression works by
>removing redundant information from a data stream. Once ORBiT's code has
>removed the redundancy, there's nothing left for the tape drive to do,
>and you incur the overhead of the compression logic and formatting
>without further reducing the amount of data on the tape.
>
>In general, I use software compression in preference to hardware
>compression, though most people recommend the opposite. The advantage to
>software compression is that it can take the type of data into account.
>My Unix backup utility, for example, knows about compressed and gzip'd
>files and does not try to compress them. On the other hand, software
>compression is practical only when the computer is not doing anything
>else during the backup, since the compression is CPU-intensive. Since my
>backups run at night with nobody else on the system, that's not an issue
>for me.
>
>-- Bruce


It would seem that another reason to do software compression is that the
backup packages that I know of include a decompressing program at the start
of the tape and can therefore be restored to any system.  Hardware
compression requires that the receiving tape drive be able to decompress the
data.


Gary Jackson
Nevada CSOS
(916) 478-6407 - voice
(916) 478-6410 - fax