Howdy, Stranger!

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


Shells Virtual Desktop
BMail.ag - Secure Email Service
Server.net
CPLicense.net
VPS Server
Buy VPN
Vultr
VMs for AI
HostDare
ReliableSite White-Label Dedicated Hosting for Resellers
InterServer VPS
BMail.ag - Secure Email Service
Best VPN
High-Performance Bare Metal Server Solutions
Karvl.com
Server Mania Cloud Hosting
DataWagon Hosting
AlphaVPS Hosting
Evoxt.com
Clouvider
VPS Hosting with NVMe
Residential IPs in the US & 4G Mobile Proxies in EU & US with Unlimited Bandwidth
ReliableSite White-Label Dedicated Hosting for Resellers
Rabisu - Hosting Solutions
Shells Virtual Desktop
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.

Can CPU steal be the guest's fault

2»

Comments

  • AvenaCloudAvenaCloud Member, Patron Provider

    @dev_vps said:

    would you mind giving me a test VPS for 3 days?
    I would like to run through my tests.

    -- I play with numbers for living

    For one day I can give you without problems, but 3 days is already too much.

  • @AvenaCloud said:

    @dev_vps said:

    would you mind giving me a test VPS for 3 days?
    I would like to run through my tests.

    -- I play with numbers for living

    For one day I can give you without problems, but 3 days is already too much.

    ok.
    I will take on that one day (24 hours) test VPS.

  • AvenaCloudAvenaCloud Member, Patron Provider

    Please create the order on the link below, but do not pay and we will activate it.
    After you create the order, please, open a ticket and mention your LET member name.

    https://secure.avenacloud.com/cart/cloud-kvm-vps-vds/

  • @AvenaCloud said:
    Please create the order on the link below, but do not pay and we will activate it.
    After you create the order, please, open a ticket and mention your LET member name.

    https://secure.avenacloud.com/cart/cloud-kvm-vps-vds/

    Thank you.
    Ticket Created! #507826

  • Seems like Avoro dedicated bullshit

  • NeoonNeoon Community Contributor, Veteran

    @szymonp said:
    Hello LET,
    my VPS is experiencing ~30% CPU steal. This has been going on for weeks. The provider claims that the node only has 45% load so it must be my fault. Is that possible?

    If the Node has 45% load, you are not going to see 30% steal.
    There is no shortage of cpu time that would explain the 30%.

    Either the node is oversold, which is very likely or your hitting the hard cpu limits.

    Thanked by 2jsg szymonp
  • AvenaCloudAvenaCloud Member, Patron Provider

    @Neoon said:

    @szymonp said:
    Hello LET,
    my VPS is experiencing ~30% CPU steal. This has been going on for weeks. The provider claims that the node only has 45% load so it must be my fault. Is that possible?

    If the Node has 45% load, you are not going to see 30% steal.
    There is no shortage of cpu time that would explain the 30%.

    Either the node is oversold, which is very likely or your hitting the hard cpu limits.

    On page 1 it is already clear what was the problem with this server. No node is overloaded.

  • @AvenaCloud said: but 3 days is already too much.

    Why? If it's VDS, he has paid for 100% of those cores.

    Thanked by 1jsg
  • Maybe you can fix this with AI. Talk to @c1vhosting
    /s

  • layer7layer7 Member, Host Rep, LIR

    @AvenaCloud said:
    For all VPSs, vCPUs are dedicated, just like memory and disk, everything is dedicated and not shared. We use KVM virtualization for VPSs, better said VDS and not OpenVZ or LXC. KVM virtualization completely separates the resources of the VPS server from other servers on the node in the cluster.
    The one I wrote about 45% was about how much the physical processors were loaded on the node.

    Hi,

    sounds like you define the "dedication" of the resources as the fact that from the perspective of the virtual server noone else ( except the customer ) can use the resources. So "inside" the rented virtual server there is only the customer and noone else.

    So you refer to the difference between a shared (web)hosting environment and your product.

    The problem is that the general understanding of VDS in context to virtual servers is that the resources are exclusive from the perspective of the hostmachine. So noone else, except the customer can use the assigned cores and ram. As to disk, this is not the case, as real dedicated disks are currently not offered usually.

    To prevent missunderstandings, maybe it wont be bad to write somewhere in your offers your definition of "dedicated" so there wont be any interpretation issues anymore.

    Thanked by 1dev_vps
  • sassliksasslik Barred
    edited October 2024

    @gbzret4d said:

    @Xrmaddness said:

    @szymonp said:

    That's sad, they even claim that the CPU core is dedicated :/

    Am I at least the daddy?

    Sorry bro, it's me..

  • dev_vpsdev_vps Member
    edited October 2024

    @AvenaCloud said:
    Please create the order on the link below, but do not pay and we will activate it.
    After you create the order, please, open a ticket and mention your LET member name.

    https://secure.avenacloud.com/cart/cloud-kvm-vps-vds/

    VPS Plan

    $29.20 USD / Annually
    CPU 2 vCores (dedicated - Not Sure)
    RAM 2 GB DDR4
    DISK 40 GB  SSD
    Data Port 100 mbps
    

    Customized YABS

    ---------------------------------------------------------------------------
     Basic System Info
    ---------------------------------------------------------------------------
     CPU Model          : Intel(R) Xeon(R) CPU E5-2683 v4 @ 2.10GHz
     CPU Cores          : 2 @ 2099.998 MHz
     CPU Cache          : 16384 KB
     AES-NI             : ✔ Enabled
     VM-x/AMD-V         : ✔ Enabled
     Total Disk         : 39.3 GB (1.8 GB Used)
     Total RAM          : 1.9 GB (256.1 MB Used)
     System uptime      : 0 days, 13 hour 21 min
     Load average       : 0.06, 0.01, 0.00
     OS                 : Debian GNU/Linux 12
     Arch               : x86_64 (64 Bit)
     Kernel             : 6.1.0-23-cloud-amd64
     Virtualization     : KVM
     TCP Control        : cubic
    ---------------------------------------------------------------------------
     Basic Network Info
    ---------------------------------------------------------------------------
     Primary Network    : IPv6
     IPv6 Access        : ✔ Online
     IPv4 Access        : ✔ Online
     ASN                : AS201670 S.C. INFOTECH-GRUP S.R.L.
     Host               : Infotechgrup
     Location           : Chisinau, Chișinău Municipality-CU, Moldova
    ---------------------------------------------------------------------------
    
     fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda1):
     ---------------------------------
     Block Size | 4k            (IOPS) | 64k           (IOPS)
       ------   | ---            ----  | ----           ----
     Read       | 96.96 MB/s   (24.2k) | 142.72 MB/s   (2.2k)
     Write      | 97.21 MB/s   (24.3k) | 143.47 MB/s   (2.2k)
     Total      | 194.18 MB/s  (48.5k) | 286.20 MB/s   (4.4k)
                |                      |
     Block Size | 512k          (IOPS) | 1m            (IOPS)
       ------   | ---            ----  | ----           ----
     Read       | 171.13 MB/s    (334) | 178.94 MB/s    (174)
     Write      | 180.23 MB/s    (352) | 190.86 MB/s    (186)
     Total      | 351.36 MB/s    (686) | 369.80 MB/s    (360)
    
     iperf3 Network Speed Tests (IPv4):
     ---------------------------------
     Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
     -----           | -----                     | ----            | ----            | ----
     Clouvider       | London, UK (10G)          | 97.5 Mbits/sec  | 95.1 Mbits/sec  | 46.8 ms
     Eranium         | Amsterdam, NL (100G)      | 97.9 Mbits/sec  | 95.2 Mbits/sec  | 42.4 ms
     Uztelecom       | Tashkent, UZ (10G)        | 94.6 Mbits/sec  | 93.2 Mbits/sec  | 95.3 ms
     Leaseweb        | Singapore, SG (10G)       | 63.0 Mbits/sec  | 87.3 Mbits/sec  | 182 ms
     Clouvider       | Los Angeles, CA, US (10G) | 84.3 Mbits/sec  | 89.9 Mbits/sec  | 176 ms
     Leaseweb        | NYC, NY, US (10G)         | 90.5 Mbits/sec  | 92.7 Mbits/sec  | 117 ms
     Edgoo           | Sao Paulo, BR (1G)        | 46.5 Mbits/sec  | 87.8 Mbits/sec  | 225 ms
    
     iperf3 Network Speed Tests (IPv6):
     ---------------------------------
     Provider        | Location (Link)           | Send Speed      | Recv Speed      | Ping
     -----           | -----                     | ----            | ----            | ----
     Clouvider       | London, UK (10G)          | 96.1 Mbits/sec  | 93.8 Mbits/sec  | 46.7 ms
     Eranium         | Amsterdam, NL (100G)      | 97.3 Mbits/sec  | 93.8 Mbits/sec  | 42.4 ms
     Uztelecom       | Tashkent, UZ (10G)        | 94.9 Mbits/sec  | 92.3 Mbits/sec  | 95.2 ms
     Leaseweb        | Singapore, SG (10G)       | 88.0 Mbits/sec  | 89.2 Mbits/sec  | 182 ms
     Clouvider       | Los Angeles, CA, US (10G) | 81.6 Mbits/sec  | 88.9 Mbits/sec  | 176 ms
     Leaseweb        | NYC, NY, US (10G)         | 91.5 Mbits/sec  | 91.2 Mbits/sec  | 117 ms
     Edgoo           | Sao Paulo, BR (1G)        | 62.7 Mbits/sec  | 86.5 Mbits/sec  | 225 ms
    
     Geekbench 5 Benchmark Test:
     ---------------------------------
     Test            | Value
                     |
     Single Core     |  684
     Multi Core      | 1320
     Full Test       | https://browser.geekbench.com/v5/cpu/23013676
    
     Geekbench 6 Benchmark Test:
     ---------------------------------
     Test            | Value
                     |
     Single Core     |  853
     Multi Core      | 1540
     Full Test       | https://browser.geekbench.com/v6/cpu/8574827
    
    ---------------------------------------------------------------------------
     Speedtest.net 
    ---------------------------------------------------------------------------
     Location         Latency     Loss    DL Speed       UP Speed       Server
    
     Dallas, US       149.37 ms   0.0%    97.51 Mbps     96.01 Mbps     i3D.net - Dallas, TX
     Houston, TX      153.12 ms   0.0%    98.79 Mbps     89.97 Mbps     Ezee Fiber - Houston, TX
     Chicago, IL      140.32 ms   0.0%    96.48 Mbps     94.38 Mbps     Enzu.com - Chicago, IL
     Miami, US        150.81 ms   N/A     98.36 Mbps     85.56 Mbps     Dish Wireless - Miami, FL
     New York, NY     120.47 ms   0.0%    96.41 Mbps     90.50 Mbps     Surfshark Ltd - New York, NY
     Miami, FL        139.04 ms   0.0%    95.19 Mbps     95.50 Mbps     ReliableSite Hosting - Miami, FL
     Spokane, WA      187.77 ms   0.0%    100.83 Mbps    81.22 Mbps     Crunchbits - Spokane, WA
     Seattle, WA      185.96 ms   0.0%    98.85 Mbps     76.29 Mbps     Misaka Network, Inc. - Seattle, WA
     Los Angeles, US  172.62 ms   0.0%    95.01 Mbps     87.24 Mbps     ReliableSite Hosting - Los Angeles, CA
     San Jose, CA     175.51 ms   0.0%    100.63 Mbps    84.64 Mbps     Misaka Network, Inc. - San Jose, CA
    
     London, UK        47.09 ms   0.0%    94.92 Mbps     97.63 Mbps     VeloxServ Communications - London
     Amsterdam, NL     43.86 ms   0.0%    96.25 Mbps     95.34 Mbps     31173 Services AB - Amsterdam
     Paris, FR         45.75 ms   N/A     96.33 Mbps     98.69 Mbps     Axione - Paris
     Frankfurt, DE     36.27 ms   0.0%    95.06 Mbps     99.41 Mbps     Clouvider Ltd - Frankfurt am Main
     Warsaw, PL        50.42 ms   0.0%    94.91 Mbps     95.49 Mbps     Play - Warszawa
     Kyiv, UA          32.43 ms   0.0%    96.42 Mbps     99.01 Mbps     O3 - Kyiv
    
     Bangalore, IN    156.99 ms   0.0%    95.17 Mbps     95.03 Mbps     Bharti Airtel Ltd - Bangalore
     Chennai, IN      206.42 ms   N/A     99.80 Mbps     83.49 Mbps     Jio - Chennai
     Mumbai, IN       143.87 ms   0.0%    98.21 Mbps     94.11 Mbps     i3D.net - Mumbai
     Mumbai, MH       213.23 ms   N/A     97.61 Mbps     82.55 Mbps     Jio - Rajkot
     Delhi, IN        168.35 ms   0.0%    96.37 Mbps     94.07 Mbps     Tata Play Fiber - New Delhi
    
     Jeddah, SA       116.59 ms   0.0%    101.26 Mbps    96.16 Mbps     Saudi Telecom Company
     Dubai, AE        161.90 ms   0.0%    104.31 Mbps    84.52 Mbps     du - Dubai
    
     Hong Kong        237.28 ms   0.0%    99.69 Mbps     35.91 Mbps     Misaka Network, Inc. - Hong Kong
     Sydney           366.61 ms   0.0%    92.72 Mbps     41.99 Mbps     Telstra - Sydney
     Perth            199.20 ms   0.0%    98.15 Mbps     58.37 Mbps     Aussie Broadband - Perth
    
    ---------------------------------------------------------------------------
     Avg DL Speed       : 97.47 Mbps
     Avg UL Speed       : 86.37 Mbps
    ---------------------------------------------------------------------------
    

    Thanked by 2Alyx 0xC7
  • VPS mentioned in this post is much better deal
    https://lowendtalk.com/discussion/comment/4050277/#Comment_4050277

    cost -- $10/year
    and performance is far better

    Thanked by 1AvenaCloud
  • AvenaCloudAvenaCloud Member, Patron Provider

    @dev_vps said:
    VPS mentioned in this post is much better deal
    https://lowendtalk.com/discussion/comment/4050277/#Comment_4050277

    cost -- $10/year
    and performance is far better

    Thank you very much for the review and I want to mention that we are now working on the storage for the node, to increase the speed 4-5 times. Maybe even more. I will come back with details when we finish the work on the storage.

  • dev_vpsdev_vps Member
    edited October 2024

    @AvenaCloud said:

    @dev_vps said:
    VPS mentioned in this post is much better deal
    https://lowendtalk.com/discussion/comment/4050277/#Comment_4050277

    cost -- $10/year
    and performance is far better

    Thank you very much for the review and I want to mention that we are now working on the storage for the node, to increase the speed 4-5 times. Maybe even more. I will come back with details when we finish the work on the storage.

    The VPS was pretty stable in windows, however two key points

    • disk iops is is pretty low and disk space is pretty low in comparison to comparable offers on LET
    • data port speed of 100 is s...l..o..w
  • jsgjsg Member, Resident Benchmarker

    @AvenaCloud said:
    For all VPSs, vCPUs are dedicated, just like memory and disk, everything is dedicated and not shared. We use KVM virtualization for VPSs, better said VDS and not OpenVZ or LXC. KVM virtualization completely separates the resources of the VPS server from other servers on the node in the cluster.
    The one I wrote about 45% was about how much the physical processors were loaded on the node.

    NO, that is not how it works. KVM vs. containers (like LXC) simply means that the guests (the stuff you sell) do have separate (well, kind of) resources that are not shared with others, so with KVM each guest has and runs their own kernel, etc. while container guests share pretty much everything.

    Note though that 'dedicated VPS' (or 'VDS') means that in addition 1 vCore ~ 1 HWT that is, a guest has a full vCore (or multiples) for themselves, while a VPS vCore might actually be 1/2, 1/4, or even less of a HWT. Hence a VDS can not experience any steal (modulo very small hypervisor hiccups) but a VPS is bound to see some steal and in particular "bleed through" of neighbour (hard) usage.

    So: Do NOT sell anything with shared cores as 'dedicated VPS' or 'VDS'!

    If number of vCores > HTWs it's not dedicated, no matter what hypervisor you use, simple as that.

  • @AvenaCloud said: Unfortunately, this whole procedure was done because of some clients who used the VPS illegally.

    Node was accessed illegally and VPS used without notice?

    @AvenaCloud said: But it is not possible to manually monitor all VPSs.

    It is possible. Question of competency.

    @AvenaCloud said: I mentioned earlier that our colleagues have already started correcting the rules and removing the limits from automations.

    Untested production env shows yet again lack of basic competency.

    @AvenaCloud said: We will deal differently with servers that are being misused.

    Will you torture them?

    Thanked by 2yoursunny nszerver
Sign In or Register to comment.