Oracle database management, tutorials, scripts, coding, programming and tips for database administrators
Oracle can exhibit 'strange' behavior with the sql_id values in V$SESSION, but just because the sql_id is NULL that doesn't mean Oracle isn't processing a long-running statement.
In Oracle releases 10.2.0.x and later join processing can be made more efficient by the use of Bloom filters, primarily to reduce traffic between parallel query slaves. Read on to learn more.
You're running processes in parallel; is it really faster? Sometimes it may not be; read on to see why parallel processing may be creating more work than you'd expect.
Week numbering schemes are interesting and getting the starting date based on that number is even more so. Read on to see what schemes there are and if you can, indeed, get the starting date you expect.
Large Oracle instances running on Linux can benefit from using hugepages. Read on to see what hugepages are and how they help Oracle run better.
Exadata is known far and wide for Smart Scans, but sometimes Oracle can do better without one. Read on to see how to know when Oracle decided to not continue with a Smart Scan.
The next big thing after Exadata is the new Oracle 12c Database In-memory feature, which will dramatically improve database performance for analytical queries and OLTP without application re-coding.
When performance suffers it's likely the DBA who gets the blame, but it may not be the database that's at fault. Read on to see what other factors can affect performance and how to identify what is causing the slowdown.
Chained and migrated rows in an Oracle database can adversely affect I/O performance, especially Exadata. Read on to discover how these rows are generated and what you may be able to do to correct the situation.
How do you ensure that the DBA doesn't, or can't, drop a table accidentally? Oracle has at least two ways to ensure that a table cannot be accidentally dropped but there are some limitations to those methods. David Fitzjarrell looks at those methods to see which one works for the DBA account.
Some business decisions may need to be redone, like making a non-unique primary key index unique. In Oracle 12c it's a simple task, but in Oracle 11.2. and earlier it's a bit more involved but still possible. Read on to see how this can be done.
Your developer has deleted a ton of data to help you free up space in the OS. Right? Wrong. Find out how to get all that free space in a table reflected all the way out to the operating system.
With a product as complex as Oracle some bugs are bound to be present. Some of these bugs are show-stoppers, and others aren't, but it does teach you to pay careful attention to the results a query delivers. Even though queries are syntactically and logically correct you can't be certain that Oracle won't do something 'behind the scenes' that can produce the wrong answer.
Do you really need an RMAN catalog to successfully recover your Oracle database? There are those that think so -- but they would be wrong. Read on to find out what you need to recover your database without a catalog.
In Release 11.2, Oracle has provided three improvements to earlier attempts at controlling parallel execution. These improvements make this feature more manageable, more scalable, and less likely to saturate server resources, such as memory and CPU, than earlier releases of the database. David Fitzjarrell discusses the first of those improvements, parallel statement queuing.