Hello experts,
We have an issue where a process runs. The process is simple and broken into two parts. First part of the process, there is a selection that happens from table JEST. The 2nd part, some logic is applied and updates happen in parallel to different tables. The longest time of this whole process is the select statement. Its as following:
SELECT
*
FROM
"JEST"
WHERE
"MANDT"=:A0 AND "OBJNR"=:A1
This select seems efficient but ofcourse, we want it to happen even faster. We sort the table from time to time, we also have some archiving in place to address the size. Still, considering this table has 2billion rows, we are looking at possibility of partitioning(possibly range partitioning on OBJNR). The table has 5 fields of which OBJNR is a key field. We are thinking, partitioning would help even further.
does this seem like the right approach?
Also, we also looked into possibility of turning on parallel processing on this table but it was not recommended by SAP as that affects DML operations. We are wondering if partitioning the table would let oracle process queries in parallel on this table by default? Can anyone confirm?