Parallel Processing and Parameters

Craig Cmehil

12/12/2017

Intermediate

15 min.

Leveraging SQLScript in Stored Procedures & User Defined Functions

You will learn

In this exercise we will modify the code of procedure get_po_header_data so that it takes full advantage of the parallel processing within HANA by using table variables.
Please note - This tutorial is based on SPS11

Step 1: Edit previous procedure

Return to your procedure called get_po_header_data.

Define two tabular output parameters which will be used to explicitly pass the results of the SELECT statements to the caller.

Step 2: Assign SELECT statements

Next, assign SELECT statements to the output parameters as shown here.

The completed code should be similar to this. If you do not wish to type this code, you can reference the solution web page at http://<hostname>:51013/workshop/admin/ui/exerciseMaster/?workshop=dev602&sub=ex2_11

PROCEDURE "dev602.procedures::get_po_header_data" (
    OUT EX_PO_CREATE_CNT TABLE(
       CREATE_CNT INTEGER,
      "HISTORY.CREATEDBY.EMPLOYEEID" NVARCHAR(10)),
    OUT EX_PO_CHANGE_CNT TABLE(
       CHANGE_CNT INTEGER,
       "HISTORY.CHANGEDBY.EMPLOYEEID"  NVARCHAR(10)) )  
	LANGUAGE SQLSCRIPT
	SQL SECURITY INVOKER
	--DEFAULT SCHEMA <default_schema_name>
	READS SQL DATA AS
BEGIN

ex_po_create_cnt =  SELECT COUNT(*) AS CREATE_CNT, "HISTORY.CREATEDBY.EMPLOYEEID"
     FROM "dev602.data::PO.Header" WHERE PURCHASEORDERID IN (
            SELECT PURCHASEORDERID
                   FROM "dev602.data::PO.Item"
      WHERE "PRODUCT.PRODUCTID" IS NOT NULL)
GROUP BY  "HISTORY.CREATEDBY.EMPLOYEEID";

ex_po_change_cnt =  SELECT COUNT(*) AS CHANGE_CNT, "HISTORY.CHANGEDBY.EMPLOYEEID"
     FROM "dev602.data::PO.Header"  WHERE PURCHASEORDERID IN (
             SELECT PURCHASEORDERID
                  FROM "dev602.data::PO.Item"
      WHERE "PRODUCT.PRODUCTID" IS NOT NULL)
    GROUP BY  "HISTORY.CHANGEDBY.EMPLOYEEID";

END

Step 3: Save and build

Save the procedure.

Perform a build on the hdb module.

Step 4: Invoke the procedure again

Return to the HRTT page and invoke the procedure again.

The CALL statement will be inserted into a new “SQL” tab

Click the Run button

Step 5: Check execution time

Check the execution time again, you may notice that it is a bit faster this time. The reason is that these SQL statements are now executed in parallel.