SUMMARY: format.dat entry for Seagate ST1480?

From: Martin Fredriksson G/TTV 35338 1036 (martinf@aom.ericsson.se)
Date: Tue Aug 25 1992 - 14:09:22 CDT


Thanks again Sun-managers!

I asked for a format.dat entry for Seagate ST1480.

The replies I received (as usually many, and informative), can
be divide into two groups:

(A) Use the SUN0424 entry already in format.dat.
(B) Use following values .....

As some of you might have guessed, the second group of replies
varied somewhat. I have compiled following table to sum it up:

--------------------------------------------------
            [1] [2] [3] [4] [5]
          ------ ------ ------ ------ ------
ncyl 1443 1490 1472 1474 1151
acyl 2 2 2 2 2
pcyl 1445 1492 1476 1476 2500
nhead 9 9 9 9 9
nsect 64 62 62 63 80
rpm 4412 3600 4412 4400 4400
bpt - 36022 37037 30132 26000
cache - - 0x11 0x14 -
trks_zone - - 9 9 9
asect - - 9 5 2
atrks - - 18 18 -
fmt_time - - - - 4
--------------------------------------------------
#blocks 831168 831420 821376 835758 828720
--------------------------------------------------

Where [1] is basically the figures provided by Seagate.
[2], [3] and [4] are "live" entries, actually in use
and working. [5] is the SUN0424 entry.

I must admit that I find all this somewhat confusing.....
More questions:

(1) First of all, the differing entries seem to indicate that
    the values are not important in themselves, but that the
    sums or products must be within limits(????).

(2) What values are really used for formating?

(3) What does "formatting" mean (I believe that this is not
    the SCSI low level formatting)?

(4) Why is the SUN0424 so different from the rest?

(5) Assuming that only the product of the values in the entries
    is important, can different combinations of values yield
    different drive performance (due, e.g., to file block
    organisation spanning cylinders)???

(6) I think that the ST1480 has different number of sectors/
    track, depending on which track (more sectors further from
    the center). If this is true, is the nsect value a mean
    value?

Sorry for being this confused. I'll appreciate ANY pointers,
hints or suggestions. I guess I can live without understanding
all this (I sure can get the disk working!), but I would feel
more comfortable knowing a bit more about it!

Thanks again,
I don't include the names of all who answered my question (you
know who you are!), as I can only access sun-managers mail on
an ASCII (non-cut-and-paste) terminal.....

Martin Fredriksson
System Manager, Ericsson Radar Electronics AB, Molndal, Sweden



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:48 CDT