IT Employment



Oracle Monitoring Best Practice

By b_sabharwal ·
Hi All,

I'm looking for some pointers towards things that DBA's need to monitor with regards to the Oracle database what are the best practiceses that a DBA team can follow / implement?

Thanks for all your help in advance.

This conversation is currently closed to new comments.

Thread display: Collapse - | Expand +

All Answers

Collapse -

space, errors, backups

by bwatkins In reply to Oracle Monitoring Best P ...

Here are a few things that I consider critical.

1. Space. Regularly check how much free space is available in tablespaces and on the operating system filesystems. Track how the free space goes down over time, so you can plan for adding more or purging data.
2. Alert log. Regularly check for the presence of ORA- errors in the alert log. Investigate any you find.
3. Backups. Monitor the running of your automatic backup scripts, whether it's RMAN or a third-party solution. You MUST have good backups. If the job fails, fix it and maybe take a manual backup if needed.

Get to know Oracle Enterprise Manager. It's the future for monitoring databases. Try to get your organization to pay for the Performance Pack, Tuning Pack, Configuration Pack add-ons so that you can start using these great tools (applies to 10g version and later).

Collapse -

Event Monitoring, Index Monitoring, Listener connectivity....

by mirbaqarali In reply to Oracle Monitoring Best P ...

Here are some more activities which DBA need to perform

1. We need to monitor events through gv$sysstat and identify the wait events.

2. Monitor indexes and rebuild unstructured indexes.

3.Ensure Oracle listener functionality .

4. Monitor Sqlnet.ora file.

5. Check and resolve in-doubt open transactions.

6. Check and resolve hung connections.

7. Refresh database (Cloning).

8. Check for snapshots/materialized views refresh and resolve issues(if any).

9. Monitor and track file system space usage for databases that have files with ?autoextend on? enabled.

10. Trim alert log and sqlnet logs every two months.

11. Maintain Buffer Cache Hit ratio above 85%.

12. Maintain Library Cache Hit ratio above 90%.

13.Maintain Dictionary Cache Hit ratio above 90%.

14. Maintain redo log space reques near zero.

15. Check CPU and Processor usage using TOP command in Linux.

16. Work closely with sysadmin to install all softwares and patches.

17. Database users security.

18. Monitoring performance.

Related Discussions

Related Forums