Here is a summary of the answers that I received to my question concerning
dump parameters for a HP35480A (mis-typed in the original post).There were
not as many responses as I had hoped for but...
As a reminder, the essential parts of the original post were:
> We have an HP3840A that I want to use for backing up several large
> partitions. Althoug the specs for the drive say that, with compression,
> one can get 5-8GB on a 90 m cassette, I am only managing to get around 4.5
> GB. The backup is being done over the net in the wee hours of the morning
> when no one is around. The dump command that I am using (under SunOS
> 4.1.x, BTW) is:
> rdump 0ubdsf 126 27000 36000 /dev/nrst0 sdxx
> The questions:
> * Are these the best parameters for this drive? I am pretty sure I have
> seen variations.
> * Can anyone supply a pointer to a good explaination of how these
> parameters are determined? It seems to me that there is a bit of voodoo
> involved. While trying to come up with the correct parameters for an
> Exabyte 8200 a couple of years ago, I was given at least 4 variations of
> block sizes, density and tape length by dealers, vendors, Sun tech support
> and other users. One thing that has always been a mystery to me is the
> tape length. The man page says it is in feet for tapes but, last I knew,
> 36000 ft != 90m.
Firstly, a couple of people thought that I was being overly optimistic in
expecting to get 5GB on a tape. This may indeed be so, call me gullible or
a victim of marketing hype.
The general consensus is, for Unix backups, one should expect a
compression ration of about 2:1. Since I got around 2.5:1, I'm doing
ok.
As for the parameters, it stills seems to lie more toward the voodoo
area. One respondent who said he had seen the dump source code said that
the parameters are multipliers (with the possible exception of the
blocksize). Another added, "There is some magic involved in
choosing these parameters so as not to overflow the internal arithmetic
of the dump program".
Unfortunately, no one supplied any sort of general formula or rule-of-thumb
for choosing the parameters based on the information that we normally
have about a tape/drive combination (i.e. length of tape in meters,
density, blocking size, etc.). Nor did anyone provide me with alternatives
to the parameters that I am now using.
I must say that I am not completely satisfied with the results of this
little survey. I would appreciate further comments on this subject
if there are any (i.e., does anyone have a magic formula or alternate
parameters that work better?).
Finally, thanks to those who took the time to answer my post. Although I
did not learn as much as I wished to, I _did_ learn.
-- David R. Steiner ISPA Research Assoc./Sys Admin University of Vechta Remote Sensing/GIS Vechta, Germany *All decisions are based on insufficient evidence.-Nickel Hunsenmeir*
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:21 CDT