DX Unified Infrastructure Management

 View Only
Expand all | Collapse all

CDM reporting over 100% total cpu

  • 1.  CDM reporting over 100% total cpu

    Posted Jun 22, 2017 11:16 AM

    UMP is showing over 100% total cpu on most of my Oracle RAC servers. Is this possible?

    I enabled debug on cdm probe to confirm collection. Is this normal?

     

    Jun 22 15:05:56:093 [140726255785728] cdm: CPU [03] - val=144.59% avg=48.75%

     

    cdm 5.71



  • 2.  Re: CDM reporting over 100% total cpu

    Posted Jun 22, 2017 07:25 PM

    There seems to be a defect in cdm version 5.7 related to CPU usage which is fixed in version 5.8 , could be related to the issue which you are facing ?

     

    5.80

    What's New:

    • Added support to monitor disk size in petabytes.
    • The probe supports monitoring within an IPv6 environment.

    Fixed Defects:

    • The probe displayed incorrect CPU usage for processors numbered 100 or more. Support case number 375031


  • 3.  Re: CDM reporting over 100% total cpu

    Posted Jun 23, 2017 10:11 AM

    Interesting, will give that a try and will post the results.



  • 4.  Re: CDM reporting over 100% total cpu

    Posted Jun 23, 2017 02:41 PM

    Unfortunately, still the same results:

     

    cdm.log:

    Jun 23 18:36:22:519 [140628169160448] cdm: CPU [03] - val=205.05% avg=69.08%

     

    This is what top shows:

     

    ~]$ top
    top - 18:39:00 up 90 days, 10:01, 2 users, load average: 4.22, 4.12, 3.47
    Tasks: 1161 total, 3 running, 1158 sleeping, 0 stopped, 0 zombie
    Cpu(s): 24.8%us, 1.8%sy, 0.0%ni, 68.2%id, 4.8%wa, 0.0%hi, 0.4%si, 0.0%st
    Mem: 198452828k total, 177456660k used, 20996168k free, 1320724k buffers
    Swap: 16777212k total, 1096952k used, 15680260k free, 17090392k cached

    PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
    37226 oracle 20 0 100g 571m 277m R 99.8 0.3 13:11.33 oracle
    30133 oracle 20 0 100g 509m 277m R 77.3 0.3 3:10.98 oracle
    30100 oracle 20 0 100g 283m 277m S 65.4 0.1 3:10.14 oracle
    29976 oracle 20 0 100g 283m 277m S 58.2 0.1 3:49.80 oracle
    29872 oracle 20 0 100g 365m 279m S 28.4 0.2 1:25.76 oracle
    29875 oracle 20 0 100g 365m 279m S 28.1 0.2 1:24.98 oracle
    29877 oracle 20 0 100g 365m 279m S 28.1 0.2 1:25.31 oracle
    29972 oracle 20 0 100g 365m 279m S 27.8 0.2 1:25.16 oracle
    55780 oracle 20 0 100g 282m 257m S 9.9 0.1 2092:08 oracle
    56058 oracle 20 0 100g 320m 255m S 7.3 0.2 4700:31 oracle
    50512 oracle 20 0 100g 318m 300m S 5.9 0.2 164:43.82 oracle
    30009 oracle 20 0 100g 283m 277m S 3.6 0.1 3:45.57 oracle
    55772 oracle 20 0 100g 336m 254m S 1.3 0.2 1775:51 oracle



  • 5.  Re: CDM reporting over 100% total cpu

    Posted Jun 25, 2017 07:07 PM

    hmmmm ok,

     

    Which version you have upgraded to?

    Does the server actually have 100 or more processors ?

     

    If the issue still persists after upgrading to latest version , may be the next best option would be talking to CA Support 



  • 6.  Re: CDM reporting over 100% total cpu

    Posted Jun 26, 2017 11:35 AM

    I upgraded to cdm 6.01.

    Only 8 processors.

    Yeah, I had a ticket opened for this issue and they were not that helpful.



  • 7.  Re: CDM reporting over 100% total cpu



  • 8.  Re: CDM reporting over 100% total cpu

    Posted Jun 26, 2017 11:31 AM

    Oops, sorry to mention it is:

     

    Red Hat Enterprise Linux Server release 6.5



  • 9.  Re: CDM reporting over 100% total cpu

    Posted Jun 28, 2017 06:04 PM

    The values shown are all from the log, what shows up in reports?

     

    For linux The CPU usage information is found by reading the file “/proc/stat”. Based on the data
    gathering interval either the 1, 5 or 15 minute average is used.

     

    That seems to indicate the 'average' value is what's used and that is in line with the results from the 'top' command.



  • 10.  Re: CDM reporting over 100% total cpu

    Posted Jul 10, 2017 11:27 AM

    In UMP, the CPU graph is showing it is over 100% as we speak. However, when I run top, I'm not seeing any values over 100%.

     

    top - 15:26:08 up 107 days, 6:48, 3 users, load average: 2.12, 4.21, 6.25
    Tasks: 956 total, 1 running, 955 sleeping, 0 stopped, 0 zombie
    Cpu(s): 0.7%us, 0.4%sy, 0.0%ni, 98.7%id, 0.1%wa, 0.0%hi, 0.0%si, 0.0%st
    Mem: 198452828k total, 186511912k used, 11940916k free, 1832692k buffers
    Swap: 16777212k total, 975624k used, 15801588k free, 32071764k cached

     

    I checked the /proc/stat file you mentioned as well:

    cpu 3832671540 470364 287495785 3156837 3488338 40453 87776382 0 0 0
    cpu0 323997031 27768 46293654 402067 14753 40409 77512994 0 0 0
    cpu1 280636160 21511 20309527 104042 115432 2 252716 0 0 0
    cpu2 282396688 21713 15397977 255751 398294 1 214243 0 0 0
    cpu3 286378960 20956 13341161 275897 162641 0 201627 0 0 0
    cpu4 368490037 30267 26874985 181318 166494 1 7272581 0 0 0
    cpu5 273919424 28533 17475039 84959 256623 0 237199 0 0 0
    cpu6 269279733 28914 14161493 122693 311160 0 211415 0 0 0
    cpu7 275539895 28369 12107646 3418 287929 0 192092 0 0 0
    cpu8 227133356 30270 28024763 162044 389809 22 350139 0 0 0
    cpu9 187368442 29767 15683882 114919 375740 5 215806 0 0 0
    cpu10 175809435 29905 12113426 385483 9956 2 192964 0 0 0
    cpu11 176258142 30294 10320821 104995 111034 1 177216 0 0 0
    cpu12 194308187 29756 16658567 275186 177204 1 210821 0 0 0
    cpu13 176042670 38456 15519030 90136 378689 1 185343 0 0 0
    cpu14 169149889 37386 12421366 283593 224453 0 173889 0 0 0
    cpu15 165963484 36493 10792439 310328 108119 0 175329 0 0 0



  • 11.  Re: CDM reporting over 100% total cpu

    Broadcom Employee
    Posted Jul 05, 2017 06:44 AM

    Hello Johnny, I see the support case 729535 is already closed. So is your question answered or are there still open questions ?

     

    Kind regards,

    Britta Hoffner

    CA Support