Post reply

Data from Old Cube to a re-designed one.

  • June 11, 2008 at 12:39 am

    #122596

    Hi,

    I am stuck up with the following situation:

    We have a huge cube partitioned at daily-level having more than 2 years of data and 4 years of partitions with partitions older than 2 years empty.

    We have re-designed the cube as part of a new release and are now working on re-building the new cube with the 2 years of data by restoring the corresponding SQLServer DB for the older data.

    Now we find some of the data missing and are stuck up how to get these missing data processed. The only place we can find the data is in our Old cube.

    My questions is, Is there a way to get the data from the old cube to the re-designed cube for these missing data?? It may not sound logical though!, as the new cube has been re-designed and does not match with the Old one. But just wanted to exhaust all the options before we really find that it is not possible.

    Any suggestions/tips??

    Thanks,

    GGKK..

  • June 11, 2008 at 5:08 am

    #827151

    I am not sure I completely understand the question. You have an old cube that has some data that is missing from your data mart and you want to extract it?

    Is the old cube Analysis Services 2005?

  • June 11, 2008 at 6:51 am

    #827203

    Old and New are AS 2000. The old cube has all 2 years of data. The data source tables (SQL SErver 2000) corresponding to the cube gets backed up leaving last 3 months data to till date at any time.

    Now that we are going to deploy the cube re-designed (there were dimension structure change and Cube schema changes), since we have to re-process the cube, the data source tables are restored from Tape in multiple stages and being processed.

    During this process we found that we are unable to find 3 months of data from backup. Now this data is present only in the old cube. Hence wanted to check if we have an option to move the data from the old cube to the re-designed one.

    Hope I am clear.

  • June 11, 2008 at 7:06 am

    #827213

    Bummer, AS2005 would be more helpful because of the drillthrough architecture.

    You are going to have to query the cube with a huge MDX statement that includes every dimension - or at least to the level in which you want to include data. Then, take the results of this query and look up the currogate keys in your new data mart to rebuild the records.

  • June 11, 2008 at 8:30 am

    #827285

    Actually, I am working on the MDX query. Since it looked to be a real huge effort as you say, wanted to check if there is any different way to do...

    Thanks much for your responses...Let me know if you have any more suggestions...

  • June 11, 2008 at 9:50 am

    #827363

    You could make a copy of the directory the data is stored in and try parsing the data file, but I think the MDX will end up being easier to do.

Viewing 6 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic. Login to reply