Howdy, Stranger!

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


Superb IO
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.

Superb IO

huluwahuluwa Member
edited March 2012 in General

OpenVZ 256M Box

root@buyvm5:~# 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, 3.98197 s, 270 MB/s

«1

Comments

  • Ash_HawkridgeAsh_Hawkridge Member
    edited March 2012

    Wow. Who's that with? I have only seen those numbers come from KiloServe :P

  • root@buyvm5 i guess BuyVM.

    Thanked by 1huluwa
  • What happens if you make count=32k?

  • @joodle said: root@buyvm5 i guess BuyVM.

    Didn't see that, i have just got my second BuyVM for our main site (DDoS filtering). Not done any tests yet, will do now though :)

  • @Damian said: What happens if you make count=32k?

    Depending on what node he's on (I found two services with that name)...

    (08:06:55) node43:~ root: dd if=/dev/zero of=test bs=64k count=32k conv=fdatasync
    32768+0 records in
    32768+0 records out
    2147483648 bytes (2.1 GB) copied, 9.43505 s, 228 MB/s
    (08:08:09) node47:~ root: dd if=/dev/zero of=test bs=64k count=32k conv=fdatasync
    32768+0 records in
    32768+0 records out
    2147483648 bytes (2.1 GB) copied, 8.21856 s, 261 MB/s
  • Nice :P What drive set up are you using @Aldryic

  • @huluwa said: root@buyvm5:~# 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, 3.98197 s, 270 MB/s

    OpenVZ eh?

    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, 9.37666 s, 115 MB/s
    

    That's on KVM

  • @VMPort said: I have only seen those numbers come from KiloServe :P

    270MB/s is low for Kiloserve.

    dd if=/dev/zero of=test bs=64k count=32k conv=fdatasync;rm -f test
    32768+0 records in
    32768+0 records out
    2147483648 bytes (2.1 GB) copied, 6.00616 s, 358 MB/s
  • @NanoG6 said: That's on KVM

    Is that with us? What node are you on, you should be getting better results than that -_-;

  • NanoG6NanoG6 Member
    edited March 2012

    @Aldryic Yes KVM04, using Virtio :-|

  • JacobJacob Member

    I know they're using Either 8 Drive RAID 10, But joel not long ago told me Fran ordered a few 12 Bay servers so I assume the new stock is on them 12 Bay servers.

    @VMPort said: Nice :P What drive set up are you using @Aldryic

  • KVM Results do always seem to come out lower from what i have seen. IE performing the test on the actual hardware node always comes back higher than within a VM.

  • JacobJacob Member

    @DotVPS Not sure, They might use them for both OpenVZ and Backup plans.

    I don't think backup plans are being released anytime soon though, not really heard anything about stock being added for them.

  • @Aldryic should I make a ticket for this?

  • @NanoG6 said: should I make a ticket for this?

    Nah, it's ok, I'll be taking a look as soon as I'm done with this audit. Had someone signup for filtered IPs just to try and DoS from those IPs, so I've got a bit of a mess to clean at the moment XD

    Thanked by 1NanoG6
  • @Aldryic said: signup for filtered IPs just to try and DoS from those IPs

    Sounds like the equivalent of spitting into the wind!

  • AsimAsim Member

    Check the awesome results I posted on WHT

    http://www.webhostingtalk.com/showpost.php?p=8035208&postcount=2

    500GB Backup VPS

    root@monster:~# dd if=/dev/zero of=iotest bs=64k count=16k conv=fdatasync
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 1.93353 s, 555 MB/s

  • AsimAsim Member

    Initially it was around 900MB/s but now its a bit down since there are other people using the same node :) obviously

  • @Damian said: Sounds like the equivalent of spitting into the wind!

    @DotVPS said: What a retard ^_^

    Heh, yup. My best guess is that they saw the IPs could take 10Gbit inbound and assumed they could push the same downstream as well. We still run our Autonull for that range, so it catches any attempted attacks. I guess they just wanted to ruin my morning >_<

    Thanked by 1djvdorp
  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @Asim said: Initially it was around 900MB/s but now its a bit down since there are other people using the same node :) obviously

    i still got some more optimizing to do to storage, but for sure that raid50 is insane.

    Francisco

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @dmmcintyre3 said: 270MB/s is low for Kiloserve.

    I'm not sure how 270MB/sec is slow, but yes, kiloserve is rockin' it for sure :)

    Francisco

  • DerekDerek Member

    cd into /dev/shm

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

    ??Profit

  • @Francisco said: I'm not sure how 270MB/sec is slow, but yes, kiloserve is rockin' it for sure :)

    I wasn't saying 270MB/s was slow, just that it's low for Kiloserve.

  • rds100rds100 Member
    edited March 2012

    @Derek only if you have 2G+ RAM though :)

  • DerekDerek Member

    @rds100 said: @Derek only if you have 2G+ RAM though :)

    :) My host loves me

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    @dmmcintyre3 said: I wasn't saying 270MB/s was slow, just that it's low for Kiloserve.

    Welp, kiloserve runs a good ship then :) I think they run 15k SAS?

    Francisco

  • InfinityInfinity Member, Host Rep

    @Francisco said: I think they run 15k SAS?

    According to their site, yes.

  • subigosubigo Member
    edited March 2012

    Why is everyone on this forum so hung up on large file single read/write speeds? This, along with "guaranteed/burst" RAM seems to be the main marketing gimmicks that get passed around as gospel here...

  • netomxnetomx Moderator, Veteran

    @subigo said: Why is everyone on this forum so hung up on large file single read/write speeds? This, along with "guaranteed/burst" RAM seems to be the main marketing gimmicks that get passed around as gospel here...

    another hate post...

  • @subigo said: Why is everyone on this forum so hung up on large file single read/write speeds? This, along with "guaranteed/burst" RAM seems to be the main marketing gimmicks that get passed around as gospel here...

    It's really the only way to realistically provide IO statistics. The higher the single read/write typically the random read/writes are fast as well.

Sign In or Register to comment.