+ Reply to Thread
Results 1 to 5 of 5
  1. Member
    Join Date
    Feb 2016
    Posts
    90

    Certifications
    MCP, MTA, MCSA - Windows Server 2012, Microsoft Specialist - Server Virtualization
    #1

    Default vSphere CPU Ready time problem

    Hello,

    I know that maybe this is not the place to ask such questions, but I have one that I could use some help with, and maybe it will also help you guys with CPU Ready Time problems.

    Let's take the following ESXi Host. 2x 12 core Intel Xeon processor and 512 GB RAM.( This doesn't really matter)

    CPU Usage, as reported by vSphere Windows Client is 9370 Mhz , Capacity is 24 x 2600 mhz, so basically both CPUs are relaxing on a beach in Bali.

    While trying to calculate the CPU Ready time on this host, for the past week, I've used this foruma :

    >>>>(CPU summation value / (<chart default update interval in seconds> * 1000)) * 100 = CPU ready %

    • Realtime: 20 seconds
    • Past Day: 5 minutes (300 seconds)
    • Past Week: 30 minutes (1800 seconds)
    • Past Month: 2 hours (7200 seconds)
    • Past Year: 1 day (86400 seconds)



    So for "Past Week", my CPU Ready Summation has an average of 173766

    So it's like this
    173766 / ( 1800 * 1000 ) *100 = 9.65 % ( WHICH IS A LOT according to VMWare ) Anything above 5% should be monitored



    >>>>BUT using this calculator :
    http://www.vmcalc.com/ the result is : 0.40 %, which is AWESOME

    I can't see the formula behind that online calculator, but which one is right??
    Reply With Quote Quote  

  2. SS -->
  3. Senior Member
    Join Date
    Dec 2015
    Location
    Quebec, Canada
    Posts
    218

    Certifications
    A+, Network+, Linux+, HP APS, VCP 3-4-5-6, VSP,VTSP, SSCP, Veeam VMCE
    #2
    I wouldnt calculate CPU Ready this way, because it is Vm specific. Also, even if your average is nice you can get a high CPU Ready if all your CPU activity is at the "same time".

    The CPU ready % is more runtime related. It is the % of time that a VM is ready to execute but there is not enough pCPU to execute.

    Use ESXtop and capture this data, if you want some "real" statistics.
    Reply With Quote Quote  

  4. Member
    Join Date
    Feb 2016
    Posts
    90

    Certifications
    MCP, MTA, MCSA - Windows Server 2012, Microsoft Specialist - Server Virtualization
    #3
    Thank you,

    I will try ESXtop and see if I have any issues with over-commitment. Thank you
    Reply With Quote Quote  

  5. Senior Member
    Join Date
    Dec 2015
    Location
    Quebec, Canada
    Posts
    218

    Certifications
    A+, Network+, Linux+, HP APS, VCP 3-4-5-6, VSP,VTSP, SSCP, Veeam VMCE
    #4
    Most of the time, vCPU are over-allocated. I have seen some SMB Windows Domain controller with 12 vCPU allocated... what a waste.. Best practice propose to begin with 1 or 2 vCPU then allocated as required.
    Reply With Quote Quote  

  6. Member
    Join Date
    Feb 2016
    Posts
    90

    Certifications
    MCP, MTA, MCSA - Windows Server 2012, Microsoft Specialist - Server Virtualization
    #5
    Thank you very much, SteveLavoie!
    I'll come back with the results of ESXtop
    Reply With Quote Quote  

+ Reply to Thread

Social Networking & Bookmarks