Maximum PC

So in the May issue story on 2TB+ hard drives hitting the wall, Gordon Mah Ung says, "Yes, you can run multiple partitions to get around the problem, but that's not optimal...."

My question is, why? What makes multiple partitions "not optimal"? I've been building my own systems for many, many years, and ever since I replaced my beloved Seagate ST4096

(check that baby out, 80MB of space stuffed into a 5.25" Full Height drive bay)

with a massive 200MB Maxtor, I've set up multiple partitions on my systems - usually separating out OS, apps, and data.

Tim

Well, reading/writing to/from both partitions on the same disk is always great fun. I only ever put multiple partitions on a disk if I plan to never exchange data between them, such as dual-booting scenarios. It's just a taboo I've had since the early millennium, when I never really understood everyone's obsession with partitioning the crap out of a drive.

But... why exactly are you still partitioning your discs like that for? Is it really just so you can have a partition, instead of a folder, with your data?

This link pretty much explains the limitations of drive's capacities over time.

As far as partitioning drives with NTFS (used by NT/2000/XP/Vista/7), I remember going to a seminar before Windows 95 and NT 4.0 were launched...

When the technician/speaker got to the section about explaining NTFS under WinNT4, I specifically remember 'performance degradation will occur if drives under size limitations are partitioned into multiple partitions.'

There's not that much difference between reading/writing to/from two partitions on the same disk and reading/writing to/from data on the same disk that I can see, assuming you partition wisely.



There's plenty of reasons to partition a disk - ease in data backup, simplified reformat/install of the OS, better performance for certain activities (OS and apps in faster portion of drive, placement of swap file away from other data, etc), and others I can't think of off the top of my head.

Tim

Yes, I was aware of why there was a problem with the big disks, but I was wondering about the reasoning on the quote I gave.



Well, 15 years is a long time to remember such a specific quote, and an even longer time to apply it to drives with limitations unthought of then, but even if that is true, I was trying to find out exactly what problems or difficulties there might be.

Do you have any idea specifically why it might not be optimal?

Thanks.

Tim

It has to do with the way the NT/2k/XP/Vista/7 read the disk. They read the entire span of the disk, from inmost - where it reads slower - to the outermost - where your read speeds are best.

When you split a drive into multiple partitions, the first partition goes to the inside, and each subsequent partition goes outboard of that. Kind of like an LP record (remember those?) in reverse.

I'll try to hunt down specifics on it for you.

I believe you have that backwards - the first partition goes on the outside and thus the faster part of the disk. This is why a process called "short-stroking" works. See http://www.tomshardware.com/reviews/short-stroking-hdd,2157.html

Tim