Order of execution plan

Below reference clearly explains the order of execution plan.

http://www.dba-oracle.com/t_order_sequence_sql_execution_explain_plans_steps.htm


Cross verify results with below examples:

Example:
Query:
select /*+ gather_plan_statistics */
e.first_name,d.department_name,j.job_title,j.job_id
from
hr.employees e,
hr.departments d,
hr.jobs j
where
e.job_id=j.job_id
and e.department_id=d.department_id
and j.job_title in ('Purchasing Clerk','Marketing Manager','Administration Assistant')
--and e.first_name in ('Sigal','Alexander','Jennifer','Susan')
;
SELECT *  FROM  TABLE(DBMS_XPLAN.DISPLAY_CURSOR(null,null,'ALLSTATS LAST')) ;

Plan:

Id	Operation	Name	Starts	E-Rows	A-Rows	A-Time	Buffers	OMem	1Mem	Used-Mem
0	SELECT STATEMENT		1		7	00:00.0	19
*  1	HASH JOIN		1	17	7	00:00.0	19	880K	880K	629K (0)
2	NESTED LOOPS		1		7	00:00.0	12
3	NESTED LOOPS		1	17	7	00:00.0	9
*  4	TABLE ACCESS FULL	JOBS	1	3	3	00:00.0	7
*  5	INDEX RANGE SCAN	EMP_JOB_IX	3	6	7	00:00.0	2
6	TABLE ACCESS BY INDEX ROWID	EMPLOYEES	7	6	7	00:00.0	3
7	TABLE ACCESS FULL	DEPARTMENTS	1	27	27	00:00.0	7

order of id execution: 4,5,3,6,2,7,1,0

It follows "Left-Right-Center" at every node from leaf to top in tree.

It follows post order traversal. Refer this.
https://www.tutorialspoint.com/python_data_structure/python_tree_traversal_algorithms.htm

Viewing the actual explain plan of Query

step1: add /*+ gather_plan_statistics */ hint in select statement as shown below.

select /*+ gather_plan_statistics */
e.first_name,d.department_name,j.job_title,j.job_id
from
hr.employees e,
hr.departments d,
hr.jobs j
where
e.job_id=j.job_id
and e.department_id=d.department_id
and j.job_title in ('Purchasing Clerk','Marketing Manager','Administration Assistant')
--and e.first_name in ('Sigal','Alexander','Jennifer','Susan')
;

step2: execute
 SELECT *  FROM  TABLE(DBMS_XPLAN.DISPLAY_CURSOR(null,null,'ALLSTATS LAST')) ;
A-rows shows actually resulted query where as E-rows shows estimated row count.

Pivot, Unpivot in Oracle SQL

Below is a sample table.

select * from olympic_medal_winners;

Pivot: Convert rows to columns.

select * from (
select noc,medal from olympic_medal_winners)
pivot (
  count(*) for medal
  in ( 'Gold' Gold, 'Silver' Silver , 'Bronze'   Bronze)

)
order by 1 desc
fetch first 6 rows only;

Note: Always maintain pivot in outer most subquery with "select * from" clause.

Below is a query which results total events,sports, gender where gold , silver, bronze is won.

select * from (
 select noc, medal, sport, event, gender
 from   olympic_medal_winners
)
pivot (
 count(distinct sport ||'#'|| event ||'#'||gender )
 for medal in ( 'Gold' gold, 'Silver' silver, 'Bronze' bronze )
)
order  by 2 desc, 3 desc, 4 desc
fetch first 5 rows only;

Note: The outer grouping will occur only if that column is not used in Pivot braces. In above case sport,gender, event columns are used in Pivot braces() so, that wouldn't be grouped in result. Only NOC column is used to group.

Using multiple measure(count, sum) and filter pivoted data.

select * from (
  select noc, medal, sport, event, gender, athlete
  from   olympic_medal_winners
)
pivot  (
  count( distinct sport ||'#'|| event ||'#'|| gender ) medals,
  count( distinct sport ) sports,
  listagg( athlete, ',') within group (order by athlete) athletes
  for medal in ( 'Gold' gold )
)

where  noc like 'D%'

Unpivot: Convert Columns to rows.

Below is an another data set representing medals count.


select * from olympic_medal_tables

Below is an example of unpivoting the data.
select * from olympic_medal_tables
unpivot (medal_count for medal_colour in (
  gold_medals as 'GOLD',
  silver_medals as 'SILVER',
  bronze_medals as 'BRONZE'
))
order  by noc
fetch  first 6 rows only;

NOC  MEDAL_COLOUR  MEDAL_COUNT  
ALG  GOLD                                  0            
ALG  BRONZE                             0            
ALG  SILVER                                2            
ARG  GOLD                                  3            
ARG  BRONZE                             0            
ARG  SILVER                                1

Note: Observe that each column is denoted with a data set name GOLD, SILVER, BROZE etc to consider each of them in one column.

Generating Calendar using Connect Prior by

Note: Replace $Start_dt, $End_dt with the actual start,end dates

SELECT

            TO_CHAR(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'Day') DAY_NAME,

            TRUNC(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM) DAY_DT,

            TO_NUMBER(TO_CHAR(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'WW') ) CAL_WEEK_NUM,

            TRUNC(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'WW') CAL_WEEK_START_DT,

            TRUNC(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'WW') + 6 CAL_WEEK_END_DT,

            TO_NUMBER(TO_CHAR(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'DDD') ) CAL_DAY_OF_YEAR,

            TRUNC(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM) - TRUNC(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'Q'

) CAL_DAY_OF_QTR,

            TO_CHAR(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'DD') CAL_DAY_OF_MONTH,

            TO_CHAR(TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - ROWNUM,'D') CAL_DAY_OF_WEEK

        FROM

            DUAL

        CONNECT BY

            ROWNUM <= TO_DATE('$End_dt','MM/DD/YYYY HH24:MI:SS') - TO_DATE('$Start_Dt','MM/DD/YYYY HH24:MI:SS')

        ORDER BY

            DAY_DT;

Gather stats of a table sytax

begin

DBMS_STATS.GATHER_TABLE_STATS('owner', 'table_name',
                                       estimate_percent=>50,
                                       block_sample=>TRUE,
                                       degree=>4) ;

end  ;

or

ANALYZE TABLE <table_name> ESTIMATE STATISTICS 1000 ROWS;

or

ANALYZE TABLE <table_name> ESTIMATE STATISTICS 50 PERCENT;

ERROR: "Internal error. The Source Qualifier [] contains an unbound field" while running a workflow in PowerCenter

Problem Description
While running a workflow in the PowerCenter, the following error message is displayed:

Internal error. The Source Qualifier [] contains an unbound field []. Contact Informatica Global Customer Support.

Cause
This issue occurs when there are one or more unconnected ports between the Source and Source Qualifier.

Solution
The error encountered is not a product bug. It can be overcome by good mapping design and usage.

SQL to find Monday and Friday of the week of a given date

select next_day (sysdate-7,'FRIDAY') Last_Friday, next_day (sysdate-7, 'MONDAY') Last_Monday from dual;