Page 1 of 4 1234 LastLast
Results 1 to 40 of 148
  1. #1

    VPS Disk I/O test

    I'm new to the VPS scene. I just bought a VPS(OpenVZ) from PhotonVPS and decided to test it first one being the disk. What can you guys make of this result?

    [[email protected] ~]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 84.3708 seconds, 12.7 MB/s

    I tested it at a different times and best I got was 17MB/s. Is this speed reasonable?

  2. #2
    Join Date
    Apr 2004
    Location
    Singapore
    Posts
    1,506
    For your reference:

    My First VPS
    RAID 1
    [svr2:/root]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 15.0042 seconds, 71.6 MB/s
    My Second VPS
    RAID-10
    [[email protected] ~]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 3.1535 seconds, 340 MB/s

  3. #3
    Join Date
    Nov 2009
    Location
    Malaysia
    Posts
    153
    Quote Originally Posted by andr3w4u View Post
    I'm new to the VPS scene. I just bought a VPS(OpenVZ) from PhotonVPS and decided to test it first one being the disk. What can you guys make of this result?

    [[email protected] ~]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 84.3708 seconds, 12.7 MB/s

    I tested it at a different times and best I got was 17MB/s. Is this speed reasonable?
    Please contact your provider, anything below 20MB/s is really unusual ..

  4. #4
    Quote Originally Posted by seikan View Post
    Please contact your provider, anything below 20MB/s is really unusual ..
    I had contacted PhotonVPS once but no difference!.

    I still get 15-20 MB/s.

  5. #5
    Join Date
    Nov 2010
    Location
    UK
    Posts
    43
    What dis they say when you contacted them. Did they awknowlwdge the problem or givd you any sort of explanation?

  6. #6
    Join Date
    Mar 2008
    Location
    Suffolk, England.
    Posts
    646
    Something's definitely not right there, I've just run this on a personal testing VPS hosted on one of our busy nodes and received the following result:


    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 5.46038 s, 197 MB/s


    I would definitely contact your provider and see what's going on!

  7. #7
    Quote Originally Posted by seikan View Post
    Please contact your provider, anything below 20MB/s is really unusual ..
    I have contacted them, they kept telling me everything's fine or that they fixed it but I always kept getting under 20 MB/s. They switched me to another node and it was worse than the last, I told them about it and they said it was fine I even asked to try another node with no answer back from them yet.

    What should I do?

    Oh also this is on a RAID 10 setup.
    Last edited by andr3w4u; 12-05-2010 at 11:20 AM.

  8. #8
    Join Date
    Oct 2010
    Posts
    152
    Interesting. Their YardVPS ($15/year OpenVZ plan) performs better than PhotonVPS plans?

    1073741824 bytes (1.1 GB) copied, 17.367 seconds, 61.8 MB/s

  9. #9
    Quote Originally Posted by skoh View Post
    Interesting. Their YardVPS ($15/year OpenVZ plan) performs better than PhotonVPS plans?

    1073741824 bytes (1.1 GB) copied, 17.367 seconds, 61.8 MB/s
    Oh wow I must be part of the unlucky bunch. It's supposed to be Raid 10 with "Enterprise Western Digital Hard Drives".

  10. #10
    Join Date
    Jun 2003
    Location
    Los Angeles, CA
    Posts
    1,506
    There is something wrong here, do you have a ticket number that I can look in to?

  11. #11
    ticket id is 608617

  12. #12
    I have a Leaf 6 plan with YardVPS and, while it worked great at first, after a couple weeks I never once saw speeds faster than 18 - 20 MB/s using the exact same dd test as you performed, and often had speeds below 15.

    I raised a ticket with them and was repeatedly ignored until I asked if upgrading to a Xen plan would offer better performance. I was told that I "should see an overall performance increase" if I did so, but that they were currently sold out and were awaiting new hardware. When I asked if there was anything that could be done in the meantime, I was told "YardVPS is our budget brand and you will not experience the best possible performance from this brand."

    At this point I zipped up all my files and moved to a new provider. It's just a matter of cancelling my account with YardVPS now. Hopefully that goes a little easier than the support process.

    PhotonVPS-Jim: If you'd like to investigate this issue, my Ticket ID number is SRT-496805. It's a shame that it had to come to me cancelling, as when the service worked, it worked really, really well.
    Last edited by ehushagen; 12-05-2010 at 09:31 PM.

  13. #13
    Join Date
    May 2007
    Posts
    1,969
    yardvps

    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 4.59436 seconds, 234 MB/s

  14. #14
    Quote Originally Posted by ehushagen View Post
    I have a Leaf 6 plan with YardVPS and, while it worked great at first, after a couple weeks I never once saw speeds faster than 18 - 20 MB/s using the exact same dd test as you performed, and often had speeds below 15.

    I raised a ticket with them and was repeatedly ignored until I asked if upgrading to a Xen plan would offer better performance. I was told that I "should see an overall performance increase" if I did so, but that they were currently sold out and were awaiting new hardware. When I asked if there was anything that could be done in the meantime, I was told "YardVPS is our budget brand and you will not experience the best possible performance from this brand."

    At this point I zipped up all my files and moved to a new provider. It's just a matter of cancelling my account with YardVPS now. Hopefully that goes a little easier than the support process.

    PhotonVPS-Jim: If you'd like to investigate this issue, my Ticket ID number is SRT-496805. It's a shame that it had to come to me cancelling, as when the service worked, it worked really, really well.
    Seems like we're in the same boat the only difference being that I'm not with their budget brand. I'm still waiting on PhotonVPS-Jim to contact me. I'm giving a chance before I consider canceling. Thanks to those that posted their disk speed test result I envy you guys .

  15. #15
    Join Date
    Sep 2010
    Location
    /usr/bin/fail
    Posts
    858
    I am also seeing slow speeds on PhotonVPS. I am on a BEAM1 plan.


    [email protected] [~]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 48.9655 seconds, 21.9 MB/s

    Even though is not a big issue as it is only used as a secondary DNS server so its low impact.


    @Jim Here is my ticket number 431828 just to help track the issue down..
    Last edited by FLDataTeK; 12-06-2010 at 12:08 AM.

  16. #16
    Join Date
    Jun 2003
    Location
    Los Angeles, CA
    Posts
    1,506
    andr3w4u -

    The issue should be resolved. There was an user spamming that was causing all IO usage. There will be occasional spike due to our system doing its backups.


    ehushagen -

    The node you were on had a degraded array that was suffering through performance issues while it rebuilds. This is unexpected but no data lost was suffered during this degradation.


    JeremyXL -

    The node you were on had cache write off due it being on a legacy system. We'll be migrating this raid through out the weeks. However currently write cache is on. However we'll be making it more optimal.



    We monitor our nodes regularly but due to IO priority on OpenVZ some will have lower priority vs those who gain a 200MB/s which are on Xen. If any one has performance issues we advise you to open a ticket with us and work with our Level 3 engineers for a resolution.

  17. #17
    Join Date
    Sep 2010
    Location
    /usr/bin/fail
    Posts
    858
    Thanks Jim.. Yes there is indeed an increase after they modified it.. Here is the same test on the same vps..

    [email protected] [~]# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 17.7395 seconds, 60.5 MB/s


    So almost 3x increase in performance..

    Thank you for your wonderful service and quick resolution..

  18. #18
    Here's mine

    DMEHosting (Denver)
    Code:
    # dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 19.7884 seconds, 54.3 MB/s
    DMEHosting (Chicago)
    Code:
    # dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 17.3874 seconds, 61.8 MB/s
    ThrustVPS (Los Angeles)
    Code:
    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 23.1792 seconds, 46.3 MB/s

  19. #19
    Join Date
    Apr 2009
    Posts
    50
    Hello

    Sorry, i'm noob and here's my result

    dd if=/dev/zero of=test bs=16k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    67108864 bytes (67 MB) copied, 4.82657 seconds, 13.9 MB/s

    Same question, is this speed reasonable? ( my vps at alienvps )

  20. #20
    Join Date
    Sep 2002
    Location
    Chicago, Illinois
    Posts
    882
    Quote Originally Posted by WebHostX View Post
    Hello

    Sorry, i'm noob and here's my result

    dd if=/dev/zero of=test bs=16k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    67108864 bytes (67 MB) copied, 4.82657 seconds, 13.9 MB/s

    Same question, is this speed reasonable? ( my vps at alienvps )
    That is a bit low, but not absolutely terrible. If your applications are working without problems then there really isn't cause for concern.

  21. #21
    Join Date
    Apr 2009
    Posts
    50
    Thanks for your advice

  22. #22
    Join Date
    Mar 2010
    Location
    Florida
    Posts
    113
    Well I just ran it on my machine and I think your numbers are way off.

    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 3.40199 s, 301 MB/s
    [email protected]:~$

  23. #23
    Join Date
    Mar 2010
    Posts
    4,487
    Quote Originally Posted by WebHostX View Post
    Hello

    Sorry, i'm noob and here's my result

    dd if=/dev/zero of=test bs=16k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    67108864 bytes (67 MB) copied, 4.82657 seconds, 13.9 MB/s

    Same question, is this speed reasonable? ( my vps at alienvps )
    Anything under 40-50 MB/s in my opinion is unreasonable. Of course multiple write speeds are more important then a single file.

  24. #24
    Code:
    [[email protected] ~]# dd if=/dev/zero of=test bs=64k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    268435456 bytes (268 MB) copied, 11.0193 seconds, 24.4 MB/s

    I am on simba. Jim, is there anything you can do?

    There is an old Ticket #602658 I had raised for the same reason.

  25. #25
    Join Date
    Mar 2009
    Location
    Austin, TX
    Posts
    934
    Quote Originally Posted by WebHostX View Post
    Hello

    Sorry, i'm noob and here's my result

    dd if=/dev/zero of=test bs=16k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    67108864 bytes (67 MB) copied, 4.82657 seconds, 13.9 MB/s

    Same question, is this speed reasonable? ( my vps at alienvps )
    I think you are ok when compared to my AlienVPS VPS...


    $ dd if=/dev/zero of=test bs=16k count=4k conv=fdatasync
    4096+0 records in
    4096+0 records out
    67108864 bytes (67 MB) copied, 8.61845 s, 7.8 MB/s

  26. #26
    Join Date
    Nov 2009
    Location
    Malaysia
    Posts
    153
    Quote Originally Posted by quad3datwork View Post
    I think you are ok when compared to my AlienVPS VPS...
    What node are you? I had same issue with AlienVPS and they moved me to another node to resolve this.

  27. #27
    Join Date
    Mar 2009
    Location
    Austin, TX
    Posts
    934
    Quote Originally Posted by seikan View Post
    What node are you? I had same issue with AlienVPS and they moved me to another node to resolve this.
    They were acquired by some company recently and they moved me to NY DC. I'm on node "ny5".

    Are you still in old DC or new DC? The old DC was worse IMHO. I can type a command and takes few seconds before it executes.

  28. #28
    Join Date
    Nov 2009
    Location
    Malaysia
    Posts
    153
    Quote Originally Posted by quad3datwork View Post
    They were acquired by some company recently and they moved me to NY DC. I'm on node "ny5".

    Are you still in old DC or new DC? The old DC was worse IMHO. I can type a command and takes few seconds before it executes.
    Ops, you are unlucky. My VPS was in ny5 last time and they moved me to ny4. Seem like their I/O issue still exists on ny5 node.

  29. #29
    Join Date
    Nov 2009
    Location
    Malaysia
    Posts
    153
    I guess my VPS with YardVPS having very inconsistent disk I/O:

    Code:
    # dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 7.78172 s, 34.5 MB/s
    
    # dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 13.3051 s, 20.2 MB/s
    
    # dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 22.8598 s, 11.7 MB/s
    
    # dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 15.8097 s, 17.0 MB/s

  30. #30
    Join Date
    Sep 2004
    Posts
    1,007
    Here's my results for Linode (NJ):

    Code:
    16384+0 records in                                                                                  
    16384+0 records out                                                                                 
    268435456 bytes (268 MB) copied, 1.16155 s, 231 MB/s
    I should note, though, that /dev/zero isn't really a valid test anymore. Some hard drives, like those using SandForce controllers, do compression to reduce write amplification. Although since they're modern SSDs, I'm not sure you could even tell, since the SATA interface might be the limiting factor at that point.

  31. #31
    Join Date
    Jun 2003
    Location
    Los Angeles, CA
    Posts
    1,506
    Depending on the plan you sit on the IO priority will be lower if you are on a lower plan. This allow us to control abuse as well as allocate resources to those who pay a bit more. A $15/ USD / year will not get you the highest priority. As you can see your values are quite consistent where you are averaging 20's - 30's. This allows us to ensure high availibility and fight abuse that might occur on our nodes.

  32. #32
    Join Date
    Mar 2006
    Location
    Australia
    Posts
    771
    I was really interested in trying out VRTServer's Cloud VPS product in comparison to ensCloud's one. VRTServer's cloud product is absolutely awful.


    [[email protected] ~]# dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 25.4469 seconds, 10.5 MB/s
    [[email protected] ~]#
    [[email protected] ~]# dd if=/dev/zero of=test bs=16k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    268435456 bytes (268 MB) copied, 40.7289 seconds, 6.6 MB/s

    Probably not even a cloud. Avoid!

  33. #33
    AlienVPS node:ny1; 11:18 a.m. here

    #####@#####:~$ dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 record dentro
    16384+0 record fuori
    1073741824 byte (1,1 GB) copiati, 60,1222 s, 17,9 MB/s
    #####@#####:~$

  34. #34
    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 102.757 s, 10.4 MB/s

    i am waiting for htop 5 sec :-) xsserver.eu

  35. #35
    Join Date
    Mar 2008
    Location
    Suffolk, England.
    Posts
    646
    Quote Originally Posted by nortol View Post
    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 102.757 s, 10.4 MB/s

    i am waiting for htop 5 sec :-) xsserver.eu
    Ouch! Which plan are you on may I ask?

  36. #36
    NL VPS BIZ

  37. #37
    I am was customer of Delimiter, but Delimiter is dead, and i has to find urgency another vps...
    Any way, i think this vps will be not the last of my vps :-)

  38. #38
    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 174.422 s, 6.2 MB/s

    I think this is my daily extra speed :-)

  39. #39
    Join Date
    Mar 2008
    Location
    Suffolk, England.
    Posts
    646
    Quote Originally Posted by nortol View Post
    dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 174.422 s, 6.2 MB/s

    I think this is my daily extra speed :-)
    Something is seriously at issue there, have you tried contacting your provider?

  40. #40
    No, i have no tried to contact with my provider, becaouse i dont care anymore. I finished with all VPs's. I have already buyed a new machine in hetzner. This is not vps, but a full server, i have now X4 from Hetzner: hetzner.de/en/hosting/produkte_rootserver/x4/

Page 1 of 4 1234 LastLast

Similar Threads

  1. Tool for stress test , check disk i/o and health of drives in array
    By turbovps in forum Colocation and Data Centers
    Replies: 2
    Last Post: 11-17-2010, 03:47 PM
  2. Replies: 0
    Last Post: 07-22-2010, 12:06 PM
  3. Replies: 15
    Last Post: 02-12-2010, 02:23 PM
  4. 1 Test VPS Only! Windows 2003 VPS - 512MB RAM / 50GB / 100GB Bandwidth
    By A Grateful Dad in forum VPS Hosting Offers
    Replies: 9
    Last Post: 08-22-2009, 09:48 PM
  5. Hard disk test 'surprises' Google.
    By encoderX in forum Computers and Peripherals
    Replies: 1
    Last Post: 02-21-2007, 03:06 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •