The latter two first. A Sequential Read operation is like reading in a large chunk (or all) of a file in one shot, from starting to end points - for example, loading an entire .exe file into RAM to be executed. In this operation the disk heads do little seeking for new data sectors because in most cases the next block to read is right after this one. On the other hand, a Randon Read operation is what happens when you start feeding out a list of individual records you want retreived from a large database file. Based on a knowledge of where in the file each record is stored, on each record retrieval the system has to tell the disk to jump from where it is to wherever the next required record is, and the heads spend a bunch of time moving to the new track location, then waiting for the disk to turn around until the right sector is being read. This is a slower process.
For some time now HDD makers have included in the unit's control board some RAM dedicated to being a buffer for this drive only. The buffers have been getting bigger as RAM got cheaper. On a read without a buffer, the drive has to position the head and then read the track there, looking at the data until the right sectors come up. Then it sends out the data requested and goes on to the next request. But with a buffer in use, the drive will read an entire track into the buffer at once - this takes one head relocation and one disk turn as before - and then it can send out the part that was requested. BUT the next request comes in and the first activity is to look in the buffer to see whether the data happens to be there already. If yes, then it is sent out VERY much faster than seeking it with mechanical movements. And in most cases this works because so many data requests are for sequences of related information. If it's not already in the buffer then we go back to the normal seek operation. So a buffer can speed up the average data access time by eliminating many of the head-seek-and-disk-turn times. By the way, a buffer also can be (and is) used for writing to the disk in bigger chunks.
A Buffered Read of sequential data is the fastest of these three. An Unbuffered Read of sequential data is next because it often eliminates the need to relocate the heads, even though there is still time taken to wait for the right disk sectors to pass under the head. And a Random Read is the slowest because it eliminates any benefit of having the buffer AND it forces a head relocation for every read requested.
Facebook
Twitter