Building a data warehouse

  • Hi guys

    having successfully clean my data and populated my dimensional and fact tables, my question is how can i insert values from the dimensional tables into the fact table

    the following are my dimensional tables:

    - Dim_Temprequset {RequestId(pk), RequestDate}

    - Dim_BankBranch {BranchId(pk), Address, town, county, postcode, telephone, typeofbranch}

    - Dim_time{ System_date, Date_in_month, day_text, day_in_week, week_in_month, month_number, month_text, quater, year}

    - Dim_session {SessionId(pk), requestId, TempId, sessiondate, sessionstart, sessionend, status, type}

    - Fact_sales {Time_key, BranchId, SessionId, temprequestId, county, branchname, town}

    can anyone help

    πŸ™‚

    New kid on the block

  • Delario (1/6/2012)


    having successfully clean my data and populated my dimensional and fact tables, my question is how can i insert values from the dimensional tables into the fact table

    the following are my dimensional tables:

    - Dim_Temprequset {RequestId(pk), RequestDate}

    - Dim_BankBranch {BranchId(pk), Address, town, county, postcode, telephone, typeofbranch}

    - Dim_time{ System_date, Date_in_month, day_text, day_in_week, week_in_month, month_number, month_text, quater, year}

    - Dim_session {SessionId(pk), requestId, TempId, sessiondate, sessionstart, sessionend, status, type}

    - Fact_sales {Time_key, BranchId, SessionId, temprequestId, county, branchname, town}

    Please note there is a "Data Warehouse" section on this forum.

    Having said that - you usually do not move data from DIM to FACT, ETL process should populate both DIM and FACT tables using the data extracted from the OLTP source/s.

    _____________________________________
    Pablo (Paul) Berzukov

    Author of Understanding Database Administration available at Amazon and other bookstores.

    Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.
  • just name them like to differentiate the tables, in reality the name convention are different

    New kid on the block

  • Delario (1/6/2012)


    just name them like to differentiate the tables, in reality the name convention are different

    :blink: would you rephrase, please?

    _____________________________________
    Pablo (Paul) Berzukov

    Author of Understanding Database Administration available at Amazon and other bookstores.

    Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.
  • ok let put it this way, I will change the name of the table

    :alien:

    New kid on the block

  • Delario (1/6/2012)


    ok let put it this way, I will change the name of the table

    Well... No matter the actual names of the tables the fact is that in a Dimensional Modeled a.k.a. star-schema Data Warehouse, by definition you end up having FACT and DIM tables and these FACT and DIM tables should be populated by the ETL process using data extracted from the OLTP source - not moving data between them.

    _____________________________________
    Pablo (Paul) Berzukov

    Author of Understanding Database Administration available at Amazon and other bookstores.

    Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.
  • got your point

    New kid on the block

  • so how can I populate the fact table

    New kid on the block

  • Delario (1/6/2012)


    so how can I populate the fact table

    Well... you designed it, don't you?

    Each row on the particular FACT table is supposed to describe a factual event on your OLTP system, isn't it? Therefore your ETL process should "E"xtract all new factual events from the OLTP system, "T"ransform them as needed to map the attributes on your FACT table then "L"oad such data into the FACT table.

    Quick example:

    You have a OLTP system that does billing for your company.

    Such a system generates Invoices - the generation of each "Invoice" is a business fact.

    On your DWH you design a FACT_INVOICE table.

    Your daily ETL process loads into FACT_INVOICE all new "invoices" generated on the OLTP system since the last time ETL ran - some related DIMension tables will get loaded alongside.

    No offense but designing a DWH is not something you do like buttering bread, you may consider to either get professional assistance or really (really, realy, really) research a lot more.

    _____________________________________
    Pablo (Paul) Berzukov

    Author of Understanding Database Administration available at Amazon and other bookstores.

    Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.
  • I understand the concept but I am having problem with the implimentation. that said all that you have said have done with clear understanding.

    New kid on the block

  • Delario (1/6/2012)


    I understand the concept but I am having problem with the implimentation. that said all that you have said have done with clear understanding.

    take a look here...this may help

    http://technet.microsoft.com/en-us/library/ms166552.aspx

    ________________________________________________________________
    you can lead a user to data....but you cannot make them think
    and remember....every day is a school day

  • Guys, it’s an Oracle Data Warehouse and questions are distributed in multiple threads...

    http://www.sqlservercentral.com/Forums/Topic1231770-391-1.aspx

    http://www.sqlservercentral.com/Forums/Topic1231314-391-1.aspx

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

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