This blog should be a quick and easy read. I will admit that this is a weird combination of topics. We are going to learn how to generate traditional, Statspack-style performance reports from the Automatic Workload Repository. This blog also contains my thoughts on the future of database tuning. And the future does NOT include administrators poring over reams of database statistics to tune Oracle databases.

I absolutely and firmly believe that ADDM and the intelligent advisors are the future of database performance analysis. In my next blog, we let Oracle do all of the tuning work for us when we generate an ADDM task from the same set of beginning and ending snapshot IDs.

It looks like I am going to have to deviate from my intended topic of discussion for one blog. Oracle has recently announced 10G Express Edition with great fanfare. I thought that I would provide you with a quick overview of the product. We’ll then get back on track next week when I discuss 10G EM performance snapshot reports.

Now that we have a good overview of 10G AWR snapshots, let’s continue our discussion by learning how to tailor the 10G AWR snapshot environment to meet our specific needs.

If you are like me, I’m sure you would rather learn how to use the performance statistics captured in the snapshots to monitor and tune your databases. But we must spend one blog learning how to configure 10G Grid Control to ensure that we have the information we need. We can’t tune if we don’t have adequate performance measurements!

Now that we have completed our discussion on the 10G EM Advisors, let’s take an in-depth look at the performance statistics capturing and reporting capabilities provided by 10G EM. Oracle 10G captures performance statistics on a periodic basis and stores the data in the Automatic Workload Repository. The individual statistic capture executions are called AWR Snapshots.

We’ll begin our discussion with a review of the 10G AWR snapshot environment. In a series of upcoming blogs will learn how to use AWR Snapshots to generate performance reports, compare two sets of performance statistics to one another, compare a captured set of statistics to a baseline and finally, how to preserve snapshot sets for future diagnostic use. When we are done with this series, you’ll be saying “Statspack, who needs THAT tired old tool anymore?"

We continue our discussion on the 10G Grid Control advisors. In this blog, I’ll provide you with a brief review of Automatic Undo Management in Oracle 10G. I’ll continue the discussion by showing you how to use the 10G Grid Control Automatic Undo Advisor to determine how much space you will need in your undo tablespace for a specified undo retention time period.

Before we return to our series of blogs on 10G EM features and functionality, I think it is important for us to discuss the 10G Enterprise Manager Standalone Java Console. Previous blogs have contained information on troubleshooting 10G EM management server to agent communications. But what happens if the entire environment goes down and you still want to use a GUI version of the tool? In addition, the management server version of 10G EM currently does not provide the capability of administering some of Oracle’s advanced features such as streams, advanced queuing, etc..

This blog will show you how to install and configure the 10G Enterprise Manager Standalone Java Console. We’ll also compare and contrast the Java Console to its big brother that uses the management server and repository.

Before we return to our regularly scheduled 10G EM topics, I'd like to provide you with my opinions on Tom Kyte's latest book.

This is the final blog of a three part series on troubleshooting 10G Grid Control agent to management server communication issues. I’ll start the discussion by summarizing a few key points from the previous two blogs.

I’ll also provide a few hints that will help you determine where to start the debugging process (management server or agent). I’ll complete this series by showing you how to activate detailed tracing on the agent and management server components. If you can’t identify the problem by analyzing error messages normally generated by the agents and management server, you’ll have to activate more detailed traces to gather additional diagnostic data.

Let's continue our discussion on debugging 10G Enterprise Manager agent to management server communications. My last blog covered some of Giant Eagle's experiences with agent administration and troubleshooting. But the management agents are only one piece of the puzzle; the problem could also be on the management server. As a result, when debugging communication failures we must also spend some time reviewing diagnostic information generated by the 10G Grid Control management server environment.

As we continue down our path of 10G Grid Control enlightenment, I thought it might be advantageous to deviate from our discussions on the advisors for yet another blog. Keeping the communications flowing between the agents and the management console can be somewhat tricky at times. I will admit that most of our issues were "self inflicted". If you are like us, you'll have to learn how to troubleshoot a problem or two until you gain experience.

We have been installing and administering the 10G agents for some time now and we think we have crested the top of the learning curve. Once you gain experience, the environment pretty works as advertised. I thought I would give you a few helpful facts on how to administer the agents and also provide you with some information on how we solved some of the problems we have encountered. In an upcoming blog I'll describe how to troubleshoot the management server.