Howdy, Stranger!

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


Accepting the challenge ! Xen SSD from 12 Eur/year ! Improved 50cents ! - Page 2
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.

Accepting the challenge ! Xen SSD from 12 Eur/year ! Improved 50cents !

2»

Comments

  • SpiritSpirit Member
    edited October 2013

    Existing Prometeus customers can apply code [temp.removed] for this plan for a fixed annual price of €12. Limited quantity !

    Fix it :p
    Temporarily removed code before too many take advantage of your mistake.

  • prometeusprometeus Member, Host Rep

    @Spirit said:
    Temporarily removed code before too many take advantage of your mistake.

    Fixed :-)

    Thanks

  • Well...

    root@xd256:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync; unlink test
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 1.06787 s, 1.0 GB/s
    root@xd256:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync; unlink test
    16384+0 records in
    16384+0 records out
    1073741824 bytes (1.1 GB) copied, 1.09922 s, 977 MB/s
    root@xd256:~#

  • MaouniqueMaounique Host Rep, Veteran

    This is good or bad ?

  • In the openvz ssd, aren't we get about 1,9 GB/s?

  • MaouniqueMaounique Host Rep, Veteran

    It varies, you can get 1.4 and 1.5 as well, however, this is Xen. Overhead is a bit higher and at such speeds is more visible.

  • @Maounique @prometeus Ah, somehow i forgot about iwstack. ;)

  • This is good or bad ?

    Really good :)

  • Buy again on DALLAS 50Cents :)

  • At last i cannot resist, got 1 50 cents plan. Should be good for personal squid proxy and openvpn server.

  • DylanDylan Member
    edited October 2013

    Is it normal for download speeds on a DALLAS XEN 128 plan to be this much slower than on a DALLASZ3? These results have been pretty consistent since I signed up for both plans.

    Xen 128

    Download speed from CacheFly: 58.6MB/s

    Download speed from Coloat, Atlanta GA: 2.61MB/s

    Download speed from Softlayer, Dallas, TX: 23.2MB/s

    Download speed from Linode, Tokyo, JP: 350KB/s

    Download speed from i3d.net, NL: 446KB/s

    Download speed from Leaseweb, Haarlem, NL: 414KB/s

    Download speed from Softlayer, Singapore: 266KB/s

    Download speed from Softlayer, Seattle, WA: 1.17MB/s

    Download speed from Softlayer, San Jose, CA: 1.17MB/s

    Download speed from Softlayer, Washington, DC: 1.35MB/s



    Z3

    Download speed from CacheFly: 108MB/s

    Download speed from Coloat, Atlanta GA: 52.4MB/s

    Download speed from Softlayer, Dallas, TX: 101MB/s

    Download speed from Linode, Tokyo, JP: 15.7MB/s

    Download speed from i3d.net, NL: 13.4MB/s

    Download speed from Leaseweb, Haarlem, NL: 18.3MB/s

    Download speed from Softlayer, Singapore: 16.4MB/s

    Download speed from Softlayer, Seattle, WA: 65.1MB/s

    Download speed from Softlayer, San Jose, CA: 65.8MB/s

    Download speed from Softlayer, Washington, DC: 66.7MB/s

  • MaouniqueMaounique Host Rep, Veteran
    edited October 2013

    This is because the buffers for network are much lower due to lower memory available so speed varies exponentially with the latency. You will have to edit sysctl.

  • prometeusprometeus Member, Host Rep
    edited October 2013

    @Dylan said:
    Is it normal for download speeds on a DALLAS XEN 128 plan to be this much slower than on a DALLASZ3? These results have been pretty consistent since I signed up for both plans.

    add the following to the /etc/sysctl.conf of you xen box

    net.core.rmem_max=16777216
    net.core.wmem_max=16777216
    net.ipv4.tcp_rmem=4096 87380 16777216
    net.ipv4.tcp_wmem=4096 65536 16777216
    
    and then run
    sysctl -p
    

    or reboot. You should see better results. You can try lower value to balance memory usage.

Sign In or Register to comment.