Hello all. I'm new here.

I just found Radified from a helpful poster on the Computing.net forums. I started a thread in the Gaming section about the partition plan I had and asked for advice on how I might improve it.

I must admit, I was quite impressed with Rad's Partition Guide. It was not too technical, but just right. And the comparison of partitions with rooms in a house was simply perfect.

The reason I'm posting this is because I was quite surprised with parts of what Rad's partition guide recommends. Parts of that seem to directly contradict what Stein Gjoen recommends in his "HOWTO: Multi Disk System Tuning" (for a Linux/Win 95/DOS system) at http://www.tldp.org/HOWTO/Multi-Disk-HOWTO.html. I used that to plan my partitions, but now I have my doubts. Admittedly, it's quite outdated and mostly intended for a multi-disk Linux multi-OS setup, but I still found it useful.

To be honest, I believe I may have misunderstood the explaination that Stein Gjoen gave of which part of the drive is which. However, I believe that the Radified partition guide could have made this clearer, too.

How about adding a diagram with LABELS? The Partition Magic picture at the top of page 1 helps a bit. But there would be no doubt if one edge of that diagram was labeled "outer/fastest" and the other end labeled "inner/slowest"? I'm guessing from the description given that the left side of the Partition Magic screen is the outer/fastest, but I'm not certain. When I read Stein Gjoen's HOWTO, I got the distinct impression that the 1st partition, that the left side of the Partition Magic screen was the inner/slowest...

Also, I'm puzzled by how the Radified partition guide states that:

"In other words [at the outer/leading/fastest edge], the drive's read/write heads won't have to travel (seek) to the far end of the drive (during normal system usage, anyway). This will provide you with a more responsive system by decreasing the drive's effective seek/access time."

The Stein Gjoen's guide clearly states that the MIDDLE of a hard drive is always the fastest! By "fastest", he takes into consideration both the Seek Time and the Sustained Transfer Rate. His guide seems to state that the outer edge has the fastest Sustained Transfer Rate because it's the edge of the platters - which are spinning past the heads at a much faster rate than at the inner/center tracks (where STR is the lowest).

The reason he gave for the middle being the fastest, is because the STR is average there and the Seek Time is at the max.

The reason he gives why the seek time is maximum at the middle is because, in real world applications, the HD heads are constantly moving and, , tend to be somewhere in the middle of the tracks - neither near the outer nor inner tracks.

If you really think about that analogy, it does make sense. Aren't the heads constantly moving during most HD seeks? If the heads happened to be caught near the outer edge when requested to find a file near the inner tracks, it would take a very long time to get there. Same way if the heads were near the inner tracks and sent to find a file near the outer edge.

But whether the heads happened to be near the inner or outer tracks, it could always reach a file near the middle faster than going all the way to the opposite side. If the heads happened to be near the middle, it could find the file near the inner or outer tracks just as quickly either way. And if the file was near the middle, it would be found very quickly indeed.

Anyway, I guess my question is: Who is right... and why? Also, is there a possiblity that hard drive technology has changed so drastically since that "Multi Disk System Tuning" Howto was written than the rules on figuring the Seek Time and STR are completely different? I'm just confused... ???

Yeah, my strategy is designed to put the most-frequently accessed files (operating system, programs, data) near the beginning of the disk, where access times are lowest (fastest) and sustained transfer rates are highest (also fastest),

.. and (as NightOwl mentioned) I also designed my strategy to keep the read/write heads in that area by placing only rarely-accessed files at the far/inner/slow end of the disk.

The center-area of the disk strategy is valid only if you have commonly-accessed files ALL OVER THE DISK.

My strategy is designed to prevent this by partitioning the disk to put the most-frequently accessed files near the beginning/outer/fastest end of the disk, and the rarely-accessed files at the far/inner/slow end.

I should also say that my knowledge of Linux is limited. I've installed a few distro's and played with them, but not much more than that.

But here's what I know: if you can keep the read/write heads near the outer/fastest end of the disk, your read/write heads won't have far to seek, so your system will respond mor quickly.

Yes, I agree you are correct that a graphic illustration would clarify. I guess it's a matter of priorities right now, with much to do and a limited amount of time .. I'm sure you understand.

I'll put it on my list, which is quite lengthy right now.