plugging a repo version number into the tabular model

  • hi we run 2019 std.  we saw this week that someone hid 3 important columns (2 are calc'd measures) from our sales cube.  i'm in the process of setting up a repo for our bim going forward.  i'm not the only one that can deploy and process.  its totally possible that these were hidden before i arrived here.

    i'm repairing the tabular model but wondering if the community has any ideas for me to easily (air tight) relate perhaps thru a pivot or dax query on the db's DATA itself a version number or date i may bake into the repo somehow?   then i can have a compare run daily that will catch if this occurs again.

    i already know that locking down security even further is an option but thats not going to happen.  And i suppose the best solution might be automatically generating a create daily that is then compared to the bim in the repo.

    BTW if my repo is of bim files, whose name is always the same ,  willl i need a uniquely named folder for each version of each different cube?

     

  • Thanks for posting your issue and hopefully someone will answer soon.

    This is an automated bump to increase visibility of your question.

  • You say that locking down security is not an option, but it should be.

    Setting up a process such that your BIM only gets deployed from a VCS is the way forward. By all means, make the deployment happen automatically on commit to branch 'x'.

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • thx phil.   may not have that option as dba's etc will sometimes do what they want.  any thoughts on verifying automatically what is running was the last bim versioned?

  • What do you want to happen in the case where the deployed version is found to be different from the VCS version?

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • probably log and email a ticket.  i (the bad cop) would start getting  the offenders to stop doing that slowly.

  • stan wrote:

    any thoughts on verifying automatically what is running was the last bim versioned?

    You can run Tabular Edit with the appropriate permissions to connect to your prod server. Then do a File open to DB and then save. Then use you favorite text compare tool.

  • stan wrote:

    any thoughts on verifying automatically what is running was the last bim versioned?

    You can run Tabular Edit with the appropriate permissions to connect to your prod server. Then do a File open to DB and then save. Then use you favorite text compare tool.

  • You could try BISM Normalizer.

    I haven't used it, but it looks promising.

    You may be able to automate the production of a schema comparison report with it.

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • thx chrissy i am assuming you are talking about this product and marking your post as an answer.    https://www.sqlbi.com/tools/tabular-editor/

  • Yes I was talking about the Tabular Editor product. Better answer would be query MDSchema_Cubes

    SELECT * FROM $system.MDSchema_Cubes

    The docs look wrong since LAST_SCHEMA_UPDATE and LAST_DATA_UPDATE have the same description.

    https://learn.microsoft.com/en-us/previous-versions/sql/sql-server-2012/ms126271(v=sql.110)

    Best answer is what Phil suggested, use VCS and a secured process.

     

  • stan wrote:

    thx chrissy i am assuming you are talking about this product and marking your post as an answer.    https://www.sqlbi.com/tools/tabular-editor/%5B/quote%5D

    Did you try BISM Normalizer? Looks like the schema compare option is more refined than a simple text compare.

    The absence of evidence is not evidence of absence
    - Martin Rees
    The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
    - Phil Parkin

  • not yet phil.   will look at it soon.

Viewing 13 posts - 1 through 12 (of 12 total)

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