Subject: | |
From: | |
Reply To: | |
Date: | Fri, 22 Aug 1997 16:28:57 +0200 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Well, my testing is done. I reported that hardware compression was
better, but in fact Stan pointed out that I had misread my own
benchmarks.
I must conclude that on a 969/200, it is better to use turbostore's
software compression over the drives hardware compression to save time.
The software compression on turbostore reports compression as follows:-
FILE DATA COMPRESSED RATIO PERCENTAGE
MEDIA 1 9913835530 2935662168 3.4:1 70%
Not bad. The largest files (60% of data) are TurboImage datasets (fairly
full).
So I guess that with 12GB before compression at 3.4:1, I could look
forward to getting about 4 times as much data on the DAT i.e. 12GB of
pre compressed data, or almost 40GB of uncompressed Data.
This is impressive stuff.
I tried to get the results of the hardware compression using SCSIDDS
utilities, but apparently, I'm not capable of deciding for myself
whether to run these diagnostics, and I must apply for a suplicen
password, so that'll have to wait.
I hope that early next year, I'll be able to report on four DDS3 drives
in parallel on a 959/400 :)
Now, if only the DDS3 drives turn out to be more reliable than the DDS2
drives...........................
Regards
Neil (see you in Chicago) Harvey
|
|
|