Hard disk partitioning is almost a must, according to what I have heard so far from my colleagues and other professionals (may be same for you). I am used to partition a hard disk at least to three drives; "C" for OS (Operating System), "D" for programs and "E for other files like documents, archives. But I got a chance to read some material on hard drive performance and it completely argued my partitioning practices.When considering the speed in reading a file from hard drive following are two important points related to partitioning.
1. Files in outer tracks are readable much faster
It is a well known theory in hard disk read/writes that performance is much higher in outer tracks compared to inner tracks.
2. When partitioning a hard disk, the first partition is created in the outer tracks
In creating partitions, the first partition is created in the outer most tracks while each new partition is created in inner tracks than the previous. So the partitions in a hard disk would look as the above diagram.
These two are important point to consider. We are creating partitions in order to organize the files according to mentally created categories; like operating system in one partition while program files in another partition. But I did not organize the files in the frequency they are used (which I should have thought of). There are some files frequently used, but placed in different drives without knowing this speed factor. For example; some java projects are stored in drive "E" which I happened to use daily.
So what are the solutions we can think of, in order to achieve the best performance of the hard disk considering these facts? I can think of following two solutions.
1. Create only one partition
One option would be to refrain from creating more than one partition and storing all the files in there. Then we would have to store the most frequently used files in outer tracks while hardly used ones in inner tracks. However this is not that possible, since user is not deciding the location of a file when saving into hard drive. The only solution will be to use a defragmenting software to organize the files later according to the accessing frequency.
2. Store files on partitions according to the frequency of use
We would create a set of partitions same as we used to do earlier, but with a different intension. Files will not be saved in each drive depending on the type or categories, but according to the frequency. The most frequently used files will be stored in drive C while least used files in Drive E. But then again, memorizing the file locations will not be easy.
What is the method are you using in your computers? Do you have any experience in above mentioned suggestions? We welcome your ideas and experiences or any other suggestions.
![[Eclipse] Access restriction: Class is not accessible due to restriction on required library](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgGWqHB7pxaPjC9UKRJMWuiVgv9sd4oi2zDdwDD10FNa9u558aPgMzByPSi9_pK1ZM5HcjkMuzIaF1iXZQvCMq80jLivRajfEzcWk4J5uIu5H9hI3AbWORSXelibbY19iaoqpUzQQLFYAUm/s72-c/Eclipse+Access+restriction+rules.png)
Hello, my name is Jack Sparrow. I'm a 50 year old self-employed Pirate from the Caribbean.
Wouldn't it be easier to just have three different drives: one for programs (around 100 GBs), the other for OS (about 20 GBs) and the third for files (as large as you need it to be). You wouldn't have to worry about memorising anymore.
ReplyDeleteWell, in the UNIX world, it's widespread for people to use logical volume managers. I first did this on Solaris, but now there's a great LVM for Linux that even has an easy GUI on Redhat and Fedora Linux.
ReplyDeleteThe advantage of an LVM is that you can merge a group of disks into a single pool of space, then you can carve out little chunks of it, and you can move and resize those chunks.
To take an example, we were running an Apache install that had about 100 web sites and about that many people putting files on it. One Friday evening, a user uploads a few big video files, fills the disk, and crashes the web server. When I found out about the problem Monday morning, I called up the sysadmins and had them add a few gigabytes to the partition.
On Monday, we talked to the guy and made him his own partition. We asked him how much space he needed, then we mounted the new partition where his directory was. At that point, he had his own pool of space that was isolated from everybody else -- so he was guaranteed storage, but couldn't mess up what other people were doing.
You might say, "Just have one big partition so you're not likely to fill it." There's some sense in that, and many Linux users are doing just that. However, our Solaris machine was connected to a Storage Area Network (SAN) where we could always get a few more gigabytes on a moments notice. If we wanted another 10 terabytes, it might have taken a few days. We could unmount a partition from one machine and remount it on another.
Systems like that are available for Windows too, but they're not quite as mature.
Hi Paul,
ReplyDeleteThanks for sharing these with our readers.
The partition picture is quite interesting. Fine. Keep it up. S. Padma. 94860 74220
ReplyDeleteIf you're using windows 7 make sure you have left enough space for system partition. Windows 7 have a famous c:\Windows\winsxs folder that grows with time.
ReplyDelete30GB might be enough, but it depends how frequently you install new programs.
Hello,
ReplyDeleteCan you kindly tell me the material in which these details are contained?
@7 - Sorry, there is no link with me. However I believe you will be able to search and read a bit to find those.
ReplyDelete