Bytes per cluster (what's so special about 512)

I’m working on a revolutionary new file system design (actually its very
much like FAT at the moment :slight_smile: but am not sure whether ‘bytes per sector’
is a hardware enforced value, or can it be specified when formatting the
partition?

Hardware vendors seem to suggest that this value cannot be changed.
But Linux (can I say that here?) allows you to specify it when formatting a
FAT partition, and Windows says it handles values up to 4096.

In the fat file system it seems quite straight-forward where this value is
stored (in the Bios Parameter Block) and so can easily be changed when
formatting the disk, so whys is 512 so ubiquitous.

Would I be way off in thinking that increasing the sector size may increase
i/o throughput when dealing with large contiguous files?

Any input would be greatly appreciated.

Darragh


You are currently subscribed to ntfsd as: $subst(‘Recip.EmailAddr’)
To unsubscribe send a blank email to leave-ntfsd-$subst(‘Recip.MemberIDChar’)@lists.osr.com

sector size is specification of the hardware devices. There are currently
two sector sizes which are found for the storage devices 512 and 1024. The
devices having sector size 1024 is manufactured by few Asian vendors and its
not very popular. All the I/0 request to media is handled in terms of
sectors, therefore the above lying fs try to optimize this fact. This they
do it by specifying cluster size of the fs in terms of sector size. I
suppose when u say the variable size of sector you mean cluster size. For
most of the disk drives you can’t modify the sector size, its factory
specified. I have seen few of the RAID disk subsystem, where you can
customize the lun created. but its again either 512 or 1024.

-----Original Message-----
From: Darragh Jones [mailto:darragh.jones@cs.tcd.ie]
Sent: Wednesday, February 20, 2002 9:38 AM
To: File Systems Developers
Subject: [ntfsd] Bytes per cluster (what’s so special about 512)

I’m working on a revolutionary new file system design (actually its very
much like FAT at the moment :slight_smile: but am not sure whether ‘bytes per sector’
is a hardware enforced value, or can it be specified when formatting the
partition?

Hardware vendors seem to suggest that this value cannot be changed.
But Linux (can I say that here?) allows you to specify it when formatting a
FAT partition, and Windows says it handles values up to 4096.

In the fat file system it seems quite straight-forward where this value is
stored (in the Bios Parameter Block) and so can easily be changed when
formatting the disk, so whys is 512 so ubiquitous.

Would I be way off in thinking that increasing the sector size may increase
i/o throughput when dealing with large contiguous files?

Any input would be greatly appreciated.

Darragh


You are currently subscribed to ntfsd as: xxxxx@Legato.COM
To unsubscribe send a blank email to leave-ntfsd-$subst(‘Recip.MemberIDChar’)@lists.osr.com


You are currently subscribed to ntfsd as: $subst(‘Recip.EmailAddr’)
To unsubscribe send a blank email to leave-ntfsd-$subst(‘Recip.MemberIDChar’)@lists.osr.com