Want to subscribe to topics you're interested in?
Become a Member

DigitalOcean Slow Disks... Redis can't handle these SSD that DigitalOcean uses

Discussion in 'Virtual Private Server (VPS) hosting' started by EckyBrazzz, Sep 19, 2019.

Tags:
  1. EckyBrazzz

    EckyBrazzz Active Member

    768
    153
    43
    Mar 28, 2018
    >>>>Click here<<<< i'm nearby......
    Ratings:
    +281
    Local Time:
    12:00 PM
    1.17.x Cluster
    10.3.x Cluster & Redis
    At one of my servers at DigitalOcean I keep getting the message:
    Code (Text):
    Asynchronous AOF fsync is taking too long (disk is busy?). Writing the AOF buffer without waiting for fsync to complete, this may slow down Redis.


    That simply means the disks are too slow. Time for https://www.ovh.com

    I see some nice installation times on the Centmin install thread.
     
  2. rdan

    rdan Well-Known Member

    4,732
    1,144
    113
    May 25, 2014
    Ratings:
    +1,708
    Local Time:
    11:00 PM
    Mainline
    10.2
  3. EckyBrazzz

    EckyBrazzz Active Member

    768
    153
    43
    Mar 28, 2018
    >>>>Click here<<<< i'm nearby......
    Ratings:
    +281
    Local Time:
    12:00 PM
    1.17.x Cluster
    10.3.x Cluster & Redis
    serverhub also with Nvme... And Amsterdam(Builds bare metal on client specifications). Guess I need some time to figure out what's the best option. I run Centmin in a Cluster. So if one preforms bad, the rest suffers from it as well. Guess I gonna mix up some companies to get the best results.
     
  4. eva2000

    eva2000 Administrator Staff Member

    42,286
    9,554
    113
    May 24, 2014
    Brisbane, Australia
    Ratings:
    +14,714
    Local Time:
    1:00 AM
    Nginx 1.17.x
    MariaDB 5.5/10.x
    test underlying disk i/o performance to be sure

    But yes not all SSD disks are created equal as are not all NVMe SSD created equal. You can have a slow NVMe be slower than an SSD and vice versa as well as bad configuration/setup or bad hardware too.
     
  5. EckyBrazzz

    EckyBrazzz Active Member

    768
    153
    43
    Mar 28, 2018
    >>>>Click here<<<< i'm nearby......
    Ratings:
    +281
    Local Time:
    12:00 PM
    1.17.x Cluster
    10.3.x Cluster & Redis
    Had a look at netdata, some up and downs. Guess it's time to go somewhere else

    Project_CMM_7300.png
     
  6. Shirkit

    Shirkit New Member

    7
    1
    3
    Feb 6, 2019
    Ratings:
    +1
    Local Time:
    12:00 PM
    1.15.8
    May I ask what application you are running? Since DO seems fine with 180-240 MB/s SSDs
     
  7. EckyBrazzz

    EckyBrazzz Active Member

    768
    153
    43
    Mar 28, 2018
    >>>>Click here<<<< i'm nearby......
    Ratings:
    +281
    Local Time:
    12:00 PM
    1.17.x Cluster
    10.3.x Cluster & Redis
    @Shirkit As they tell you at DigitalOcean "Happy Coding"..... And you will notice it. Even with the expensive RAID 10 as they tell.

    I had the same issue with AMAZON AWS due the fact their EBS block storage I/O is too slow. But that was in 2013/14 or so.