DX Unified Infrastructure Management

Expand all | Collapse all

Space issue in CA UIM

  • 1.  Space issue in CA UIM

    Posted May 02, 2017 06:08 AM

    Hi Guys,

     

    In CA UIM, i am facing memory space issue. In last 2 months CA UIM has consumed more than 500GB data, also server which was having 200gb space is running out of space.

    Most of the data in server is storing in

    hub>q>health_index_alarm

    hub>q>health_index_QOS_message.

     

    I have tried the data retention also but it was also not working.



  • 2.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 02, 2017 06:14 AM
      |   view attached

    Try restarting the health_index probe.

    Look in the health index log for details of the error. Might need to increase the log level to 3 or maybe 5.

    Are you using the health_index data? If not then you could just empty both the health index queues ( in the hub gui).

    That will give you the space back, but if the queues start to build again then you will need to investigate why the health index probe isn’t draining the queues.

    Cheers

    Rowan



  • 3.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 02, 2017 09:55 AM

    It is worth noting that, largely because of issues like this, and the fact that a poll of customers could not find anyone who was actually using the data, the health_index probe has been removed from UIM 8.5.1.  So if you're able to upgrade to that version, this problem will no longer exist.



  • 4.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 07, 2017 11:29 PM

    Hello, nicaknice.

    If health_index probe is not there in your Primary HUB Robot, you are able to turn off these queues.

     

    health_index_alarm

    health_index_QOS_message

     

    Hope it helps.

    Yu Ishitani



  • 5.  Re: Space issue in CA UIM

    Posted May 08, 2017 09:32 AM

    I want  data from health_index_alarm and health_index_QOS_messgae to create SLM reports but surprising point is more than million of data is in queue which is filling Primary hub though i have increased the size of PRIMARY still data is not loading into DB.



  • 6.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 08, 2017 08:02 PM

    Please confirm

     

    - health_index probe is up running ?

    - The number in column "Sent" of the queue (in HUB probe GUI > Status) is increasing while the number in column "Queued" is also increasing ?

     

    Regards,

    Yu Ishitani



  • 7.  Re: Space issue in CA UIM

    Posted May 09, 2017 07:52 AM

    1. we are using version 8.5.1, so their is no health_index probe, only from the hub probe now we can check  and uncheck the health_index qos... 

     

    2. Yes, count is increasing only in column queue, not in sent..

     

    We don't have any space issue in db and DB is up and running.



  • 8.  Re: Space issue in CA UIM

    Posted May 09, 2017 07:55 AM

    Also please give me an insight on "baseline_engine.BASELINE_CONFIG" and "baseline_engine.Qos_messgae", i am experiencing a same Queue issue in these two as well.



  • 9.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 11, 2017 04:44 AM

    Hello.

     

    health_index_* queue must be disabled in UIM 8.5.1 since no probe is reading these queues.

     

    baseline_engine.* queue is read by baseline_engine probe, therefore please confirm if the probe is up.

     

    Regards,

    Yu Ishitani



  • 10.  Re: Space issue in CA UIM

    Posted May 12, 2017 09:33 AM

    Hello,

     

    1.If i will disable this queue from where i can get historical data to create SLA reports of Availability and of other things.I know there is a QOS for  availability with a name of QOS_SERVICE_AVAILABILTY for servers, is their any co-relation between Health index qos and availability qos?

     

    2. Yes baseline_engine probe is up and running, still data is in queue.



  • 11.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 15, 2017 06:15 AM

    Hello.

    There is no relation in between health index qos and your Availability QoS.

    Please deactivate and activate baseline_engine probe and check the queue.

     

    Regards,

    Yu Ishitani



  • 12.  Re: Space issue in CA UIM

    Posted May 15, 2017 06:22 AM

    Hi Yu Ishitani,

     

    I have already reactivation and restart thing multiple times, but still lot of data is in queue and filling the hub server space.

    Snapshot for your reference.

     



  • 13.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 15, 2017 06:35 AM

    Increase Memory Size setting in baseline_engine probe in baseline_engine.cfg.

    <startup>/<opt>/java_mem_max

    <startup>/<opt>/java_mem_init

     

    Regards,

    Yu



  • 14.  Re: Space issue in CA UIM

    Posted May 15, 2017 06:53 AM

    I am using this configuration.

    Error which i am getting is 

    Incorrect format for Ids

    Would you please tell me what is this above error?

     



  • 15.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 15, 2017 10:59 PM

    This looks a problem.

    Would you please open a Support case and engage support engineer for troubleshooting.

     

    Regards,

    Yu



  • 16.  Re: Space issue in CA UIM

    Posted May 16, 2017 04:05 AM

    Hi Yu,

     

    I did some tweaks in heap memory and now things are working fine. Thanks for continuous support.



  • 17.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 16, 2017 04:12 AM

    Glad to know its working now.



  • 18.  Re: Space issue in CA UIM

    Posted May 16, 2017 05:41 AM

    But do I need to make some tweaks in DB side as well because in last 5 months, my DB has been filled more than 800GB. 



  • 19.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 16, 2017 10:40 PM

    DB space issue is much bigger topic than we handle in the community post.

    Consult DB vendor and DBA is best path.

     

    From product side, what we can control is data retention period only.

     

    Regards,

    Yu



  • 20.  Re: Space issue in CA UIM

    Posted May 29, 2017 05:01 AM

    what should be the standard retention policy for a large infra, where we have approx 1400 devices and some standard applications.



  • 21.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 29, 2017 08:32 PM

    The data retention policy is not something we suggest, but it is up to you.

    Here is default setting in data_engine probe.

     

    Raw Data - 6 months

    Historical Data - 2 years

     

    Regards,

    Yu



  • 22.  Re: Space issue in CA UIM

    Posted May 30, 2017 04:28 AM

    Thanks Yu



  • 23.  Re: Space issue in CA UIM

    Posted May 30, 2017 06:16 AM

    Hi Yu,

     

    When we set raw data as 6 months and historical as 2 years, what will be our performance data period. I mean will be able to visualise performance data (say in report) for the 6 months raw data or the 2 years historical data? Can you clarify? Even I am having issues in our database where our raw data is 6 months and historical is just 1 day. Due to this, our maintenance jobs are getting failed almost all week and we want to find out an alternate way.

     

    -kag



  • 24.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 30, 2017 06:30 AM

    When reaching the end of RAW data retention day, raw data is rolled up to hourly average, day average, then we call them as historical data.

    So as long as I understand, retention period of historical data should be larger than raw data.

     

    Regards,

    Yu



  • 25.  Re: Space issue in CA UIM

    Posted May 30, 2017 06:53 AM

    Hi Yu,

     

    In that case for the 6 months/2 years data, will we be able to pull performance reports for the 2 years period.

     

    Also, what is the sole purpose of the historical data. I have few other questions here and if you can point me to a good documentation that details all of this, then it will really benefit me.

     

    -kag



  • 26.  Re: Space issue in CA UIM

    Broadcom Employee
    Posted May 31, 2017 04:51 AM