[ The PC Guide | Systems and Components Reference Guide | Hard
Disk Drives | Hard Disk Logical Structures and File Systems | Partitioning, Partition Sizes and Drive Lettering ]
Special-Purpose Partitions and Other Partitioning Issues
There are some circumstances in which you will want to set up partitions that are
smaller in size than usual, or where you might want to dedicate a partition to a specific
use, or ensure that it occupies a particular place on the disk. Here are some of the
issues you may want to take into account when considering how to partition your disks,
aside from the frequently-discussed matter of slack:
- Partition Placement: Most hard disks use zoned
bit recording, which means they hold more data per track at the outermost edge of the
disk than they do at the innermost edge. As a result, the outer tracks tend to deliver
better performance than the inner tracks do. Since the outer tracks are used first, this
means that the first partition on a physical disk volume will be slightly faster than
subsequent ones. If you have certain files that require higher performance than others,
placing them in a partition at the beginning of the disk is preferred.
- Dedicated Partitions: Notwithstanding my long
argument against splitting the disk into many partitions where each is for one type of
file, there are special situations where it may make sense to dedicate a partition to one
use. The most common case where a partition is dedicated to a specific use is for the
virtual memory swap file for a multitasking operating system. This file is very important
since it is used often for certain types of heavy processing, and being able to control
the exact properties and location of the partition that it uses can be advantageous. Then
again, this is really an optimization, and of less importance in newer systems than older
ones.
- Cluster Sizes for Special-Purpose Partitions: Again, having specific partitions
for certain types of files can cost you flexibility, but in some cases it can make sense.
If you are doing a lot of work with large multimedia files, you may want to intentionally
bump up the cluster size to a larger value, or at the very least, not worry about making
the partition small in order to avoid large cluster sizes. There is less overhead when
using larger clusters--doing a sequential read of a 10 MB file on a volume that uses 32
kiB clusters means 319 "next cluster" lookups in the FAT. Reading this entire
file on a volume with 2 kiB clusters increases this to 5,119 lookups. Another issue is
that since every cluster is a contiguous block on the disk, having a larger cluster size
means a greater percentage of the file is in continuous blocks--less fragmentation. This
means better performance for long sequential accesses. Frequent defragmentation of a disk
with smaller clusters will mitigate this effect, but using larger clusters is easier.
- File System Utility Performance: With modern large hard disks, enormous
partitions are possible under FAT32. Partitions with tens of gigabytes and hundreds of
thousands of files can take a very long time to process by disk scanning utilities,
defragmenters and the like. Using smaller partitions cuts down on the time required to
perform these tasks. If you have some data that changes frequently, and other data that is
mostly static, separating them onto different partitions may make sense, since you can
defragment the static files less frequently.
The points above mean that the ideal place in many cases for the swap file under
Windows, for example, is in a dedicated partition at the start of the second disk in a
two-disk system, and this is what I have set up myself in the past. On one of my PCs, I
had two disks of approximately the same size and speed, and the swap file was arranged to
take up the entire contents of the first partition of my second hard disk, about 63 MiB. I
used Partition Magic to set this partition's cluster size to 32 kiB, even though a
partition of this size would normally only use 2 kiB clusters. See here for more on swap file optimization.
Today, I don't generally bother with such minor optimizations. All PC hardware is much
faster than it was, and I don't really see much need to optimize to this degree. On a
system with a single hard disk and a modern operating system, you may be better off to
just leave the swap file on the C: drive, especially if you defragment regularly. Modern
defragmenters will optimize the swap file in a special way, moving it to the start of the
partition where transfer performance is the best. Of course, you can still use a dedicated
swap file partition if you want to, and some people still like to do this.
 Next: Partition Type Conversion
Home - Search
- Topics - Up
| |
