Topic
  • 6 replies
  • Latest Post - ‏2013-09-02T12:44:34Z by MrBhatt
MYMJ_Glen_Kroon
MYMJ_Glen_Kroon
3 Posts

Pinned topic Problem when analyzing raw nmon data

‏2013-08-07T11:52:03Z |

Peeps,

I'm having an issue when trying to create the nmon data using the analyzer 3.4a for my VIO servers. The VIO server uses EMC powerpath and there is a significant amount of powerpath devices. I'm sampling the data every 5 secs. It errors when analyzing the EMCBSIZE. This the error I get:

The specified dimension is not valid for the current chart type

 

When I open up the debugger in visual basic it goes to thisbreak point:

Chart1.Chart.ChartWizard Source:=DiskData, Gallery:=xlLine, Format:=2, _
    Title:=ChartTitle, CategoryLabels:=1, SeriesLabels:=1, HasLegend:=True

 

Any ideas, help? Happy to provide more information if needed. Thanks.

 

 

 

Attachments

Updated on 2013-08-07T12:04:30Z at 2013-08-07T12:04:30Z by MYMJ_Glen_Kroon
  • MrBhatt
    MrBhatt
    11 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-08-18T15:48:11Z  

    You may wish to try with earlier version is nmon (3.3/3.2)......

    And yes 5 sec sampling rate is too close an interval, you must be having huge files (>~30MBs) for each day.... Other than generating unexpected excel errors this may kill your PC capability to allow to work on something else in parallel

     

    Consider decreasing the interval to 1 min, unless you are not trying to make any specific observation..

     

     

    Please share the file if possible for a closer look...... and and try on my 4GB machine ;-)

     

     

  • MYMJ_Glen_Kroon
    MYMJ_Glen_Kroon
    3 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-08-19T04:09:19Z  
    • MrBhatt
    • ‏2013-08-18T15:48:11Z

    You may wish to try with earlier version is nmon (3.3/3.2)......

    And yes 5 sec sampling rate is too close an interval, you must be having huge files (>~30MBs) for each day.... Other than generating unexpected excel errors this may kill your PC capability to allow to work on something else in parallel

     

    Consider decreasing the interval to 1 min, unless you are not trying to make any specific observation..

     

     

    Please share the file if possible for a closer look...... and and try on my 4GB machine ;-)

     

     

    Hi Lokesh,

    Thanks for your response. I've sent you a copy of the raw nmon data as requested in an email. Please have a go and let me know how it progresses. Cheers.

  • MrBhatt
    MrBhatt
    11 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-08-19T10:05:29Z  

    Hi Lokesh,

    Thanks for your response. I've sent you a copy of the raw nmon data as requested in an email. Please have a go and let me know how it progresses. Cheers.

    Hi Glen,

    When i tried parsing the file using nmon3.4, i seeing some error.  Unfortunately nmon3.3e/g are not loading on my PC (64 bit OS).

    My TP is under going some updates currently, hopefully it should be able to load it afterwards.
        
    In the meantime I parsed the data with a home made script and could see good chart around CPU data . From this is could see,
        -it has generated all 180 intervals which were targeted by nmon command.
        -The sampling interval with the current data is 1 min, which you said problem is 5 sec with a problem file.

     

     

    Consider following snippet from nmon analyzer, this looks specific to you (EMC powerpath) situation...

    from nmon man page

     -l         Specifies the number of disks to be listed on each line. By default, 150 disks are listed per line. For EMC disks, specify a value of 64.
     

    from nmon_analyser readme

    1. The value of the -l flag controls the number of hdisks per sheet on the DISK sheets and per line on the BBBD sheet.   There are two factors to consider when choosing this value.  Excel has a limit of 256 columns per sheet; however, both NMON and NMON_Analyser use some columns, so the upper limit is really 250.   The second factor is that Excel VBA has an upper limit of 2048 bytes for input line length.   This particularly affects users of EMC systems that use long hdisk names (e.g. hdiskpower123).   The default of 150 is safe for such systems.   Other users may set the value to 250 in order to reduce the number of output sheets.
  • MYMJ_Glen_Kroon
    MYMJ_Glen_Kroon
    3 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-08-20T04:47:58Z  
    • MrBhatt
    • ‏2013-08-19T10:05:29Z

    Hi Glen,

    When i tried parsing the file using nmon3.4, i seeing some error.  Unfortunately nmon3.3e/g are not loading on my PC (64 bit OS).

    My TP is under going some updates currently, hopefully it should be able to load it afterwards.
        
    In the meantime I parsed the data with a home made script and could see good chart around CPU data . From this is could see,
        -it has generated all 180 intervals which were targeted by nmon command.
        -The sampling interval with the current data is 1 min, which you said problem is 5 sec with a problem file.

     

     

    Consider following snippet from nmon analyzer, this looks specific to you (EMC powerpath) situation...

    from nmon man page

     -l         Specifies the number of disks to be listed on each line. By default, 150 disks are listed per line. For EMC disks, specify a value of 64.
     

    from nmon_analyser readme

    1. The value of the -l flag controls the number of hdisks per sheet on the DISK sheets and per line on the BBBD sheet.   There are two factors to consider when choosing this value.  Excel has a limit of 256 columns per sheet; however, both NMON and NMON_Analyser use some columns, so the upper limit is really 250.   The second factor is that Excel VBA has an upper limit of 2048 bytes for input line length.   This particularly affects users of EMC systems that use long hdisk names (e.g. hdiskpower123).   The default of 150 is safe for such systems.   Other users may set the value to 250 in order to reduce the number of output sheets.

    Hi Lokesh,

    I'm so sorry to have misled the forum and you - it wasn't 5 secs. That was a typo - I was actually sampling at 60 sec intervals and I have now changed the nmon to be invoked with the -l 64 option. I will see how that goes tomorrow. Today nmon data is too late. Thanks again for the response and your time looking into this. Much appreciated.

    Kind regards.

  • MrBhatt
    MrBhatt
    11 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-08-20T09:00:51Z  

    Hi Lokesh,

    I'm so sorry to have misled the forum and you - it wasn't 5 secs. That was a typo - I was actually sampling at 60 sec intervals and I have now changed the nmon to be invoked with the -l 64 option. I will see how that goes tomorrow. Today nmon data is too late. Thanks again for the response and your time looking into this. Much appreciated.

    Kind regards.

    Not an issue at all Glen,

    It would  be interesting to see how well this option suits the current purpose and also ensuring that any desired piece of information has not been trimmed out.

    I would be eagerly waiting to see the file......

     

     

    Tnks

    ......

    MrBhatt

     

     

     

  • MrBhatt
    MrBhatt
    11 Posts

    Re: Problem when analyzing raw nmon data

    ‏2013-09-02T12:44:34Z  

    Hi Lokesh,

    I'm so sorry to have misled the forum and you - it wasn't 5 secs. That was a typo - I was actually sampling at 60 sec intervals and I have now changed the nmon to be invoked with the -l 64 option. I will see how that goes tomorrow. Today nmon data is too late. Thanks again for the response and your time looking into this. Much appreciated.

    Kind regards.

    Hi Glen,

    Interested in knowing if "-I 64" was of any help  for you ?

     

    Tnks

    ......

    MrBhatt