Add dummy Like condition '%%' inorder to use index scan

Question

Hi,
I have a table like below , which have data which capture each minute(columnB) transactions
CREATE TABLE tableA
columnA     CHAR(6),
columnB     Date,
columnC     Number(3,0),
CRT_ID    CHAR(8),
CRT_TS    TIMESTAMP(6),
UPD_ID    CHAR(8),
UPD_TS    TIMESTAMP(6),
CONSTRAINT PK_tableA_colAB PRIMARY KEY (columnA , columnB )
When I query the table, to get the list of transaction between particular date, it goes for TABLE FULL SCAN.( cost and execution time also high)
Total no of records in the table(tableA) is 13669094, the below query returns only around 150 to 200 no of records
select columnA,columnB,columnC,Crt_Id,Crt_ts
from tableA
where columnB between TO_DATE('06/28/2013','MM/DD/yyyy') and to_date('06/29/2013','MM/DD/yyyy')
when i use the query like below it will use the INDEX SCAN (cost and execution time also less)
select      columnA,columnB,columnC,Crt_Id,Crt_ts
from         tableA
where      columnB between TO_DATE('06/28/2013','MM/DD/yyyy') and to_date('06/29/2013','MM/DD/yyyy')
and           columnA like '%%'
Please advice is it good to add condition like '%%' inorder to use the index scan. Also kindly let me know if it works in the same way in all environments.

Answer 1

Hi RP0428,
Thank you very much for your response.
1. Are you collecting statistics on the table and indexes?
2. Post the exact command you use to collect those statistics.
Recently my DBA had gathered the statistics on 17-Dec-2013 22:01:32(LAST_ANALYSED). I am unaware about what query he ran to gather the statistics.
After that it executed in 28 seconds itself, before it took several minutes.
My concerns is only gathering the statistics periodically will improve performance.
Table is growing with thousands of records daily. Already it have 13669094 records. In order to avoid Full Table Scan
From my knowledge, I feel creating composite Range-Hash partitioning will helpful in improving the performance.
3. Post info about the data distribution for the two columns used by the index. This is 'counts and amounts' for each of the two columns and a GROUP BY date showing the skew of the values.
Plz find the details below
ColumnA
ColumnB
AA1
118800
AA2
117600
AA3
118200
AA4
118200
AA5
118200
AA6
118800
AA7
118800
AA8
117600
AA9
117600
AA10
117600
AA11
118200
AA12
118200
AA13
37234
AA14
118200
AA118
18450
AA119
96600
AA120
105000
AA121
105000
AA122
105600