1

I have two server, one with a SSD, one with a HDD. I've been running hdparm to see the differences, but the results for Timing cached reads and Timing are quite equal/HDD is faster:

SSD:

Timing cached reads:        26080 MB in  2.00 seconds = 13053.64 MB/sec
Timing buffered disk reads: 1558 MB in   3.00 seconds = 518.77 MB/sec

Timing O_DIRECT cached reads:   966 MB in  2.00 seconds = 482.99 MB/sec
Timing O_DIRECT disk reads:    1528 MB in  3.00 seconds = 508.71 MB/sec

HDD:

 Timing cached reads:        28248 MB in  2.00 seconds = 14137.88 MB/sec
 Timing buffered disk reads: 496 MB in    3.01 seconds = 164.78 MB/sec

 Timing O_DIRECT cached reads:   994 MB in  2.00 seconds = 496.24 MB/sec
 Timing O_DIRECT disk reads:     486 MB in  3.01 seconds = 161.55 MB/sec

The buffered disk reads are way faster on the SSD, but I'm unsure if I should worry about above mentioned performance? How could I improve this if necessary/possible?

baao
  • 113
  • 8
  • 2
    Aren't cached reads from memory? (Not from disk at all?) That's awfully fast for even an SSD and extremely unlikely with a platter. Or, do you have a large array? – Aaron Copley Dec 22 '14 at 20:26
  • 2
    `hdparm` is not a proper disk benchmarking tool. Use something better e.g. See http://serverfault.com/q/295516/37681 – HBruijn Dec 22 '14 at 20:27

1 Answers1

6

The cached reads aren't from the underlying device. They aren't representative at all of the performance of the different devices, neither of which could come anywhere near that kind of throughput. Ignore that number.

Shane Madden
  • 112,982
  • 12
  • 174
  • 248