Publishing data

Publish messages for existing application data (initial load):

  1. Start the Publish Data (ttdpm5205m000) session.
  2. Select the All Data Sets option or select all data sets that start with LNA under the Specific Data Set option.
  3. Select the Publish Initial Load option for all companies, packages, modules and tables.
  4. Optionally, select the Count option and click Publish. This will show the number of tables and records, which will give an indication of the time required to publish LN’s application data. If required, change the company and table selection to break up the publication run.
  5. Select the Publish option, set the Limited to field to 0 Records (means all records), and set the Message Size field to 5 Mega Byte.
  6. Click Publish.
  7. Inspect the log file and error file, using the View buttons.
    If you see publish errors such as “Currency rate not defined between currency EUR and home currency USD for rate type INT on date 15-Mar-2017” you are missing a currency rate for the data warehouse currency and exchange rate type. Add and approve currency rates in the Currency Rates (tcmcs0108m000) session. For the above error message, use currency base USD, currency EUR, exchange rate type INT and an effective date before 15-Mar-2017.

    A known error for the production order table tisfc001 is: “Error in lead-time calculation in Workweek of Availability Type STD.”

    LN publishes standard calculated fields for lead time between dates according to the company calendar. If the lead time cannot be calculated, the above error is given. Typically, this can be resolved by starting the Calendar Code (tcccp0110m000) session, increase the calendar end date to beyond the dates used in tisfc001, update the calendar, and republish the data.

If you have executed an initial load for a product and you add another product, the added data sets contain tables that have already been published. You can execute an initial load for all tables in the added data sets. Alternatively, you can compare data sets and skip the initial load for tables that have already been published. This can save publishing time, storage space in Data Lake, and processing time in Birst.