PCBUILD Archives

Personal Computer Hardware discussion List

PCBUILD@LISTSERV.ICORS.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
David Gillett <[log in to unmask]>
Reply To:
PCBUILD - Personal Computer Hardware discussion List <[log in to unmask]>
Date:
Thu, 15 Oct 1998 19:46:52 -0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (72 lines)
On 14 Oct 98 at 18:23, Tom Turak wrote:

> you may want to reconsider putting the swap file in a small D:
> partition.

> Recording methods differ from drive to drive, but when
> a partition is going to be so small it is only 5% of the disk capacity,
> you can reasonably assume it is going to consist of adjacent
> cylinders.

  I have never, to my knowledge, seen a partition that did not
consist of contiguous cylinders.  I'm not sure what Tom is suggesting
here.

> Outside cylinders closest to the boot sector are read at a higher
> transfer rate than inside cylinders on most drives, because some
> recording methods allow more sectors per track in those cylinders,
> so more bytes are read with each revolution.

  When making a general recommendation, the difference between "some"
and "most" may be crucial.  It's true -- a drive that uses a
recording method that puts more sectors on outer tracks will show
faster transfer rates on data stored there, and some drives show
performance characteristics which strongly suggest that they do this.

> Putting the swap disk on the inside cylinders cancels a lot of the
> advantage in putting the swap disk in its own partition.

  At very least, you want the swap file to be permanent and
contiguous, and creating it in a dedicated partition is a simple way
to make sure the latter is achieved.  For absolutely best results,
you want it on a separate drive from your application data, on a
separate controller if at all possible.  Locating it on the fastest
part of the main drive -- assuming that it is one where this makes
any difference -- is, I think, less important.  [This is clearly
somewhere where opinions may differ.]
  Since you can have multiple primary partitions on a drive (I
believe the limit is 4), there is nothing to stop you having a small
first "swap" partition and a second, much larger "work" partition; as
long as the second is the one marked "active", and booted from, it
will be known as "C:" and so the other will be assigned the letter
"D:".  [Extending DOS's rules for assigning drive letters across a
mix of primary and extended partitions across multiple drives can get
confusing, but two primary partitions on a single drive can only be
enumerated two ways, and if you give them descriptive volume labels,
you should be ready for any problems with them.]


> As for step by step instructions, Mark Rode covered it all in his
> post. One thing to consider, you can partition and format a drive
> on the secondary channel to have a primary dos partition and an
> extended partition, as long as you don't install any software
> before the drive is moved to the Master on the primary channel,
> otherwise all your drive letters will be shuffled, causing trouble
> with your registry and such.

  There is one gotcha with this approach -- Microsoft's FDISK knows
that you can only boot from drive 0, so it will not let you mark a
partition active while the drive is on the secondary channel.
  There are two work-arounds.  The simple/cheap one is to have a
bootable floppy, so that after you install the drive as a primary
master, you can get to it to run FDISK and set the partition you want
to boot from to be active.
  The other alternative is to use some third-party tool that is more
accomodating -- I like Partition Magic for this.

David G

                                  -----
                PCBUILD mailing list -  http://nospin.com
         Bob Wright:[log in to unmask] - Drew Dunn:[log in to unmask]

ATOM RSS1 RSS2