Topic
  • 4 replies
  • Latest Post - ‏2012-04-13T15:00:19Z by SystemAdmin
bhead
bhead
25 Posts

Pinned topic nmon analyser 3.2.3

‏2006-11-27T20:50:16Z |
Recently upgraded 3 servers to nmon v11e, analyser 3.2.3 and nmon2rrd 10.5. MEMNEW chart produced by nmon2rrd and rrdtool only shows proc memory usage, system and FScache are not showing up. Also noticed two DISK READ charts and two DISK WRITE charts, first chart shows the scale up to "8.0 P", second chart up to "140 T", same set of disk drives.

Tried to download the nmon output and run through the analyser to compare, kicked out a >65k lines must be pre-sorted error, so I sorted the file with sort -A, now getting "Run-time error 62 input past end of file". Using Excel 2002 SP2.

AIX version is 5.3 ML03, mix of hardware, 671, 595, 590 LPARS. The ouput I'm having trouble with the analyser has a lot of samples, 700 at 120 seconds, 67,842 lines. I know it's recommended to stay in the range of 300-500 but users are pushing for 2 minute samples. If it's not possible then I'll push back.

I apologize if this has already been asked and answered, I searched through the forum for some time and haven't had any luck so far. Thanks in advance.
Updated on 2012-04-13T15:00:19Z at 2012-04-13T15:00:19Z by SystemAdmin
  • SystemAdmin
    SystemAdmin
    2402 Posts

    Re: nmon analyser 3.2.3

    ‏2006-11-27T21:09:34Z  
    I thought I'd resolved all the errors with large files but maybe not. If you want to send the orginal, zipped, .nmon file to steve_atkins at uk.ibm.com I'll take a look at it.
  • bhead
    bhead
    25 Posts

    Re: nmon analyser 3.2.3

    ‏2006-11-28T17:42:30Z  
    I thought I'd resolved all the errors with large files but maybe not. If you want to send the orginal, zipped, .nmon file to steve_atkins at uk.ibm.com I'll take a look at it.
    In offline correspondence we've determined the problem with the analyser was self inflicted, mea culpa. In my defense I'm having a problem with my ftp client in "auto" mode, thought it was downloading in ASCII mode as opposed to Binary. So if you get a "Run-time error 62 input past end of file" using the analyser against a large pre-sorted file make sure you download it using ASCII mode.

    The problem I'm seeing with the MEMNEW graph out of the nmon2rrd still exists, only displays proc % of memory, missing system and cache. I'm taking a large number of samples (701) which isn't recommended so I'm going change it to something more reasonable and see what happens.

    Thanks to Steve and Nigel for all their assistance.

    > I thought I'd resolved all the errors with large
    > files but maybe not. If you want to send the
    > orginal, zipped, .nmon file to steve_atkins at
    > uk.ibm.com I'll take a look at it.
  • Pradeep@ACN
    Pradeep@ACN
    1 Post

    Re: nmon analyser 3.2.3

    ‏2012-04-13T14:53:31Z  
    Hi all

    I also seem to have the same issue

    Run time error '62' Input past end of file when running the file through nmon analyser:
    nmon Analyser version: V3.3.g 04-May-11

    I am using Office Excel 2010 (32 bit) on Windows 7.
    The nmon file is tracking 40 CPUs and sampling every 30 seconds 1400 times (to cover our entire batch window). The nmon file is 4MB.
    Running on AIX 6.1

    Would really appreciate some assistance on this.

    Pradeep
  • SystemAdmin
    SystemAdmin
    2402 Posts

    Re: nmon analyser 3.2.3

    ‏2012-04-13T15:00:19Z  
    Hi all

    I also seem to have the same issue

    Run time error '62' Input past end of file when running the file through nmon analyser:
    nmon Analyser version: V3.3.g 04-May-11

    I am using Office Excel 2010 (32 bit) on Windows 7.
    The nmon file is tracking 40 CPUs and sampling every 30 seconds 1400 times (to cover our entire batch window). The nmon file is 4MB.
    Running on AIX 6.1

    Would really appreciate some assistance on this.

    Pradeep
    send the file to steve_atkins at uk.ibm.com and I'll take a look