Knowledgebase
Knowledgebase
Portal Home > Knowledgebase > Articles Database > SATA vs SSD a real life experiment with 42000+ rows
SATA vs SSD a real life experiment with 42000+ rows
Posted by bdwebservices, 10-22-2016, 04:40 PM Hello People, Yesterday I did an experiment with SATA and SSD hosting. I stored 42,742 rows in MySQL and I wrote same php/mysql code. I used Chrome https://chrome.google.com/webstore/d...pmkogoaohcdbig or Firefox https://addons.mozilla.org/en-US/fir...ded-statusbar/ or Firebug to see calculate actual loading time. SATA : https://360workspace.com/test.php (bdwebservices.com) SSD : http://104.152.168.33/~bdwebser/test.php (Crocweb.com) SSD : http://162.219.248.245/~bdwebser/test.php (Eleven2.com) Seems SSD 10 to 15 Seconds faster than SATA.
Posted by TonyB, 10-22-2016, 05:43 PM This isn't an accurate test as there are no guarantees you're in fact hitting the disk on each request. With the use of i/o caches the first read after a write could be read from memory or even a subsequent read of the data. You also need to factor in how the data ended up being stored on the drives. If I write the entire table out immediately and don't write any other data it could be a completely sequential read due to lack of fragmentation. You really cannot compare SSD vs mechanical disks and use a testing methodology like this on a web host. There are all kinds of underlying technologies being utilized that could cause the disks to be irrelevant in the test.
Posted by baremetalservers, 10-24-2016, 03:45 AM Have you ever tested the PCIE STORAGE`s (https://en.wikipedia.org/wiki/NVM_Express)? It would be so great to see the results between this device and SSD.
Posted by SenseiSteve, 10-24-2016, 01:38 PM While I agree with @TonyB, I still like to see the results of these types of tests here. I think as long as we accept it for what it is, that's fine.
Add to Favourites 
Print this Article
Also Read
