DX Infrastructure Management

Expand all | Collapse all

How to monitor performance counters of linux server?

  • 1.  How to monitor performance counters of linux server?

    Posted 03-01-2019 09:15 AM

    Hi All,

     

    We have an urgent requirement to monitor performance counters of few Linux servers . I tried using Cdm probe but it is not covering all required counters.

    For windows i have used nt_perf probe and it is fulfilling the requirement but for Linux i am not able to find any probe.

    Please let me know if anyone has done this.

    PFB details of few counters required to be monitored:

    Processor utilization (For each core and cosolidated) 
      %User Time o    %Processor Time o    Interrupts / Sec  
    Memory utilization
     o    Available Bytes o    Committed Bytes o    Pages /Sec o    Page Faults/sec o    Page Reads/sec o    Page Write/sec o    Cache Faults/sec o   Swap In o   Swap Out  
    Disk utilization
    o    % Disk Read Time o    % Disk Write Time o    Disk Read bytes/sec o    Disk Write bytes/sec o    Avg Disk Read Queue Length o    Avg Disk Write Queue Length

     

     

    Thanks

    Anmol



  • 2.  Re: How to monitor performance counters of linux server?

    Posted 03-01-2019 10:49 AM

    rsp can get some of it.

    rsp Metrics - CA Unified Infrastructure Management Probes - CA Technologies Documentation 

    but

    linux admins are typically big on scripting (not me),

    so something like a script that runs every five minutes,

    the script runs the commands to get the above values,

    if the value is > *** an output file is written to with the violation

    logmon is setup to read the output file

    logmon and a different profile for each type of violation so an alarm specific to the problem will be generated

    if qos is needed

    then it is uim 9.0.2 & restmon

    write a json file to be used in creating a custom rest probe to do the needed monitoring