Question

Can this query below be optimized?

select
     max(date), sysdate - max(date)
 from
     table;

Query execution time ~5.7 seconds

I have another approach

select
    date, sysdate - date
from
    (select * from table order by date desc)
where
    rownum = 1;

Query execution ~7.9 seconds

In this particular case, table has around 17,000,000 entries.

Is there a more optimal way to rewrite this?

Update: Well, I tried the hint a few of you suggested in a database development, although with a smaller subset than the original (approximately 1,000,000 records). Without the index the queries runs slower than with the index.

The first query, without index: ~0.56 secs, with index: ~0.2 secs. The second query, without index: ~0.41 secs, with index: ~0.005 secs. (This surprised me, I thought the first query would run faster than the second, maybe it’s more suitable for smaller set of records).

I suggested to DBA this solution and he will change the table structure to accommodate this, and then i will test it with the actual data. Thanks

Answer 1

Is there an index on the date column?