Topic
5 replies Latest Post - ‏2012-05-03T10:32:11Z by YuvanK
YuvanK
YuvanK
8 Posts
ACCEPTED ANSWER

Pinned topic Archiving Cube Data

‏2012-04-30T05:09:53Z |
Hi!

Wanted to know how exactly cognos archives data from powercubes.

Firstly let me explain my situation. Our client wants to open powercube packages in powerplay and view historical data. Our position is that each day our cubes are refreshed each day with the current days data. But our client wants to be able to view data from the cube that exists one, two or even weeks before the present date. They would also prefer to store this historical cubes in a separate file system.

So therefore, my question is: how does cognos handle archiving of data from powercubes, can this even be done?
Thanks!
Updated on 2012-05-03T10:32:11Z at 2012-05-03T10:32:11Z by YuvanK
  • SystemAdmin
    SystemAdmin
    15496 Posts
    ACCEPTED ANSWER

    Re: Archiving Cube Data

    ‏2012-04-30T05:25:00Z  in response to YuvanK
    Powercubes are stored in .mdc files, that you can copy and archive just as any file, all the data is in there.
    You can add aditional datasources in Cognos Connection and point it to archived .mdcs to browse older cubes.

    It's good practice to store a few months of old cubes just in case. Usually you just add a step “copy a cube with timestamp” in cube build script.
    • YuvanK
      YuvanK
      8 Posts
      ACCEPTED ANSWER

      Re: Archiving Cube Data

      ‏2012-04-30T06:45:29Z  in response to SystemAdmin
      Many Thanks ykud!

      but if you could elaborate on your suggestion of using a script to automate the cube archiving process. How would i do this and where would it be done? what type of script?

      Also, would this mean then i would be able to generate separate cubes for each date range and save them externally?

      Thanks!
      • SystemAdmin
        SystemAdmin
        15496 Posts
        ACCEPTED ANSWER

        Re: Archiving Cube Data

        ‏2012-04-30T11:50:35Z  in response to YuvanK
        You can run cube build with a batch file (syntax depends on your Cognos version), here are some sample threads on topic:
        http://www.ibm.com/developerworks/forums/thread.jspa?threadID=279337
        http://www.ibm.com/developerworks/forums/thread.jspa?threadID=401918
        or
        http://www-01.ibm.com/support/docview.wss?uid=swg21389316
        http://publib.boulder.ibm.com/infocenter/c8bi/v8r4m0/index.jsp?topic=/com.ibm.swg.im.cognos.ug_cogtr.8.4.0.doc/ug_cogtr_id11786CommandLineOptions.html

        Regarding date ranges: that can be achieved by time-partitioning the cube
        http://publib.boulder.ibm.com/infocenter/c8bi/v8r4m0/topic/com.ibm.swg.im.cognos.ug_cogtr.8.4.0.doc/ug_cogtr_id7761DefiningTBPCube.html

        “I wanted to know if it is possible to create a new connection string apart from the primary connection string for the current data cubes and make a separate one for all the archived data. Can i modify the connection setting so that i may be able to point to the older cubes when my users want to view reports from historical data?”

        Yes, you can create another connection string pointing to older cubes. You can point it to any cube you want. If you want to switch reports, you'll have to create a FM package with parametrised datasource, pointing either to old or current cubes and pass that parameter in reports.
        • YuvanK
          YuvanK
          8 Posts
          ACCEPTED ANSWER

          Re: Archiving Cube Data

          ‏2012-05-03T10:32:11Z  in response to SystemAdmin
          Thanks so much ykud, you've been very helpful!

          But with your suggestion of using a Parameterized Connection string in FM. How exactly is this done, is there a link you could provide that explains this?

          Very much Appreciated!
    • YuvanK
      YuvanK
      8 Posts
      ACCEPTED ANSWER

      Re: Archiving Cube Data

      ‏2012-04-30T08:05:30Z  in response to SystemAdmin
      Also, one other thing...

      I wanted to know if it is possible to create a new connection string apart from the primary connection string for the current data cubes and make a separate one for all the archived data. Can i modify the connection setting so that i may be able to point to the older cubes when my users want to view reports from historical data?

      thanks