Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


What tests do I need to report slow disk speeds to tech support?
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

What tests do I need to report slow disk speeds to tech support?

rchurchrchurch Member

What tests do I need to report slow disk speeds to tech support?

Read speeds have dropped very low, and write speeds are not that good, but passable. I need tests for both.

Comments

  • 4n0nx4n0nx Member

    I searched for fdatasync and copied the first result

    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync

  • vfusevfuse Member, Host Rep

    You can use iostat / iotop to monitor things.

    Thanked by 1rchurch
  • @vfuse said:
    You can use iostat / iotop to monitor things.

    Can iostat and iotop generate a report or a graph for a given time period?

    For the mean time I am interested in an equivalent of the dd write speed test, which can give info on how slow read speeds are.

  • vfusevfuse Member, Host Rep

    @rchurch said:
    For the mean time I am interested in an equivalent of the dd write speed test, which can give info on how slow read speeds are.

    You can create a cronjob:

            • root iotop -botqqq --iter=3 >> /var/log/iotop
    Thanked by 2XiNiX vladka24
  • ClouviderClouvider Member, Patron Provider

    iotop shows I/O usage aggregated per process. How does this prove the disk is slow? Max it can prove that some process is hogging the disk, if any.

    DD is a good idea but no graph. Have you considered simply doing one test and reporting to support?

  • NyrNyr Community Contributor, Veteran

    If your server is really slow and you still need to prove it for the support to fix it, there is something seriously wrong.

  • jlayjlay Member
    edited June 2015

    You might want to consider testing using hdparm.

    For example:
    [root@db1 ~]# hdparm -Tt /dev/xvda

    /dev/xvda:
    Timing cached reads: 17114 MB in 2.00 seconds = 8569.12 MB/sec
    Timing buffered disk reads: 918 MB in 3.00 seconds = 305.59 MB/sec

    Thanked by 1rchurch
  • how do you know disk rw is low if relevant tests have not been performed?

Sign In or Register to comment.