In the Database pattern the drive is processing a stream of requests to read and write 8KB random-address data blocks. The ratio of read to write requests is changing from 0% to 100% with a step of 10% throughout the test while the request queue depth varies from 1 to 256.
You can click these links to view the tabled results:
We will build diagrams for request queue depths of 1, 16 and 256.
This test is usually boring at a request queue of 1, but this time we’ve got a lot of interesting facts to note.
The Caviar Green S8 indeed betrays a total lack of deferred writing, its performance being just terrible compared to the others.
The Samsung F4EG looks most depressing under unaligned load, too. Its performance is limited by its read speed, so its results are low at high percentages of writes.
Samsung’s F2 and, surprisingly, F4 are rather slow, too.
And finally, we can see the WD drives behave in very different ways. The server-oriented version is better under mixed loads than its 1.5-terabyte desktop cousin. The new Caviar Black Y6 falls behind its predecessor at mixed loads.
When the request queue gets longer, the 1.5-terabyte drives from WD go ahead. They are just perfect for server applications. The difference between them (and between the Caviar Black Y6 and E3 for that matter) is negligible.
The Hitachi drives are somewhat better than the Samsung ones, but the new firmware doesn’t bring any special benefits for them.
The SpinPoint F4 is the best in the Samsung camp but its advantage over the low-speed models is very small.
Once again we can note two drives with very poor performance: the Caviar Green S8 with no deferred writing and the Samsung F4EG under unaligned load.
We’ve got the same standings here as at the smaller queue depth.