Brother-Eagle® DB2 Performance
Advice: Avg Result Size

Avg Result Size < 10

OLTP databases tend to work with small result set sizes. As an example, consider online shopping sites that you may be familiar with. How many items do you add to your cart? How many addresses do you have on file with the vendor? How many credit card numbers does the vendor store on your behalf?

As a pretty good rule of thumb, if the average result set size ( ROWS_SELECTED / SELECT_SQL_STMTS ) is less than ten, the monitored database is an OLTP database.

Avg Result Size >= 10

If the average result set size is greater than, or equal to, ten, then the monitored database has data retrieval characteristics typical of a Data Warehouse database. If you believe you have an OLTP database and the average result set size is greater than ten, then you have got quite the problem child database on your hands.

• Brother-Eagle will indicate a warning if the average result set size is uncommonly large.
• Brother-Eagle will indicate an alert if the average result set size is outrageously large.
• If you observe either a warning or an alert, "make friends" with your End Users who are running queries that report large sums of data. Some End Users "ask for the world" and then scan their reports looking for the information they really want. You may be able to help these End Users write better queries, or identify staff who should attend IDUG for some SQL education.

---

It is important to make the distinction between OLTP and Data Warehouse since this distinction should influence your monitoring and tuning approaches. Throughout the Expert Advice for Brother-Eagle, you will notice that several performance metrics provide two sets of advice: one for OLTP and another for Data Warehouse.

---

View full list of DB2 LUW advice topics