Handling High Number Records in ABAP Table

Question

Hi,
Other than creating indexes is there other option to handle in an efficient way searches in an ABAP table containing more than 30 million records?
Thanks in advance.

Answer 1

Hi Ruben,
for me indexing would be the way to go: It totally depends on your access pattern. If you often read by a primary key and have no big volume queries around it would be ok.
mentioned physical Partitioning is a good option IF you can find a meaningful partition key that more or less evenly spreads the data over the partitions (i.e.  12 FISCPER that divides the 30 Mio. into partitions of 2- 3 Mio records) and that this key is used by all the quries you would start on that table. The index would be equi-partitioned. You have your DBA involved for this option; SE14 supports the physical partitioning of a table.
the other options mentioned (OPEN CURSOR...) are not much of help if you have to do number crunching;
(however, to write good SQL is the key to good performance so avoid the common pitfalls mentioned there).
You may consider a reliable archiving strategy that keeps your table on a certain row level while you move the "old" data in a kind of history table from time to time. If you need all data you would UNION  both tables.
you can hint an SQL with a PARALLEL clause to speed up the processing (i.e. you retrieve all or a big amount of the 30 Mio rows)
bye
yk