Subject: Re: newfs (and disklabel) question
To: None <>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: port-alpha
Date: 05/10/2003 16:09:05
> sectors/track: 63
> track/cylinder: 255
> sectors/cylinder: 16065 (1)
> total cylinders: 521
> total sectors: 8380080 (2)

> (1) is indeed the product of the previous two numbers, but (1) *
> total cylinders gives 8369865, not (2).  What is going on?

It is almost certainly a ZBR disk, for which the actual sectors/track
value varies depending on cylinder number.  The total sector count is
probably accurate; the reason the multiplication doesn't come out right
is that one of the three values is probably calculated by dividing the
total sector count by the other two and discarding the fraction.

Since 63 and 255 are each one less than a pwoer of two, and since
floor(8380080/(63*255)) is 521 (and floor(8380080/(521*63))=255, and
floor(8380080/(521*255))=63), I feel quite confident this is exactly
what's going on.  Especially since it is..unlikely..that the disk
actually has as many as 128 platters (255 trk/cyl, but you can in
principle use both sides of each platter) or as few as 63 sectors per

> Can I just set all these values arbitrarily?

In my experience you can, but I have comparatively little alpha
experience.  I long ago started using 64 s/t and 32 t/c for my SCSI
disks, which gives me "cylinders" that are, conveniently and not
coincidentally, exactly 1M.  Whether this would work for you depends on
many things, most of which I don't know.  I did check, and my alpha's
disks are using such settings, but I've just the one alpha, and SRM
versions may affect this if you intend to make it your boot drive.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B