The Method R Workbench software package adds the incredible data mining and data manipulation capabilities of Method R Tools to the Method R Profiler, our response time profiling tool, and Method R Trace, our zero-click performance data collector for Oracle SQL Developer. Method R Trace gives you the easiest way there is to collect Oracle trace files. The Profiler gives you the story your trace file has to tell; a complete account of response time for the end-user experience you’ve traced. Method R Tools gives you flexibility and versatility. Method R Tools gives you a language for data mining trace files with sorting, grouping, and filtering features. With Method R Tools, you can query thousands of trace files in one command. It helps you adjust trace files that need fixing.

No other software comes close to the precision and trustworthiness of Method R software.
Alex Gorbachev
Chief Technical Officer
The Pythian Group

Buy yours now

You Need a Profiler

A profile is a report of exactly how your system spends your user’s time. It’s how you answer simple-sounding business questions like these:

  • That click takes 20 seconds now. It used to take .5. Why?
  • That report took 2 hours. Could it run in 10 minutes?

As simple as these questions sound, you can’t answer them with the tools you use today. That’s why managing performance is way more difficult for you than it needs to be.

Profiling is what DBAs and sysadmins do in production when a program takes longer than we want, or when we just wonder if it could go faster. Profiling is what developers with good code hygiene do before they hand their code to the next person in the software lifecycle, to make sure their code is as tight and fast as possible—to make sure they didn’t do anything inefficient that might be particularly embarrassing under load.

We’ve always had trace files, but most developers do not know how to read them. The Profiler makes things very easy and gives me the evidence I need to demonstrate the problem is in the code, not the database.
Dorrie Keyes
DBA Manager
Indianapolis, Indiana

Sometimes, a Profiler is Just Not Enough

Sometimes, even when you have the best Oracle profiling software in the world, you still find yourself opening up that trace file to have a look? Why? There are a number of reasons:

  • The time scope is wrong on your trace file. You’re trying to diagnose an 11-second business task execution, but your trace file contains 48 seconds’ worth of data, so you need to crop your trace file. But cropping a trace file is hard. It’s not something you do with a text editor. It’s something you can do by deleting lines with a text editor.
  • You need to drill down in a dimension that your Profiler doesn’t drill. For instance, maybe you need to see your read call durations grouped by I/O size.
  • Your trace data is scattered across more than one file. Maybe your system uses a connection pooling architecture. Maybe you’re just fighting a long-running PX operation.
  • You’re looking for the region of trace data lines when something took way longer than you expected, but which lines are the ones you’re looking for? And once you find them, how are you going to convert that 13-digit Oracle Database tim value to a wall clock time, so you can join your information back to your Oracle Database Active Session History data?
Method R Tools are basically an all-in-one ETL + BI solution for Oracle database trace data.
Alex Gorbachev
Chief Technical Officer
The Pythian Group

Power. More Power

It’s hard to describe how good it feels to have to have utter domination over your performance data. Imagine knowing, with one command, which of your thousands of trace files has the best (well, worst) example of a latch contention problem in it. Imagine knowing with just one command: can you actually incur a library cache miss on an exec call? If you have just a little instrumentation in your application, you can answer questions that long-time managers of Oracle systems will find utterly amazing, like, which IP address is experiencing the worst “book order” response times?

Within half an hour, I presented to my colleague, the offending SQL statement. He needed almost 2 days to figure out the same, after long discussions with other team members, fiddling with report options in tkprof and the trace analyzers and trying to see it all in context; regarding the info presented in his reports.
Marco Gralike
Principal Database Consultant
AMIS

Be More Valuable

The best thing about understanding your trace data? Because trace data directly describes your end users’ response time experience, it’s the perfect data for understanding your users and your developers. When a user comes to you with, “My click took 23 seconds,” your response is, “Yes, I see that,” instead of, “I just don’t get it; my dials are all green...”

Because trace data describes response time in terms of database and operating system calls, it's the perfect data for discussing your problem with your DBAs and your system administrators; even your network and SAN admins. When your SAN administrator tells you, “But your average read latencies are awesome,” your response is, “But our users are spending more than 13 hours each day waiting for single-block reads that take longer than 250 µsec apiece.”

There were some unanticipated benefits from the Profiler. One is the gratification of actually helping a co-worker. She was so enthusiastic she even took the Intensive SQL Tuning class. We got her a Profiler license and she’s been busy reworking the VB app. I don’t know her very well since I’ve been in this job for a short time, but it felt so good to help someone who had been working on the same thing for so long that it had become a chore to come to work. She has more time now and is working on porting the app to Apex. Again, a little training on the method and the Profiler has gone a long way to improve her morale and productivity. You don’t see that in the marketing literature.
Tony Aponte
Database Administrator
Clearwater, Florida

Buy yours now

What’s in the Package

When you buy the Method R Workbench, here’s what you get:

Method R Software Information Library

Check out our library for more information about Method R software:

The Fine Print

Here are the definitions of the Method R Software Licensing terms. Don’t hesitate to contact us if you have any questions.

Per-Person License. A Software Per-Person License restricts the number of people who can run the Software. Each Software Per-Person License unit entitles one specific person to run the specified Software on any system, using any login id. A Per-Person license is not transferrable to another person. With a Per-Person license, the distribution contains the licensed software for all the platforms we support.

Per-Instance License. A Software Per-Instance License restricts the number of Oracle Instances with which the Software can run. Each Software Per-Instance License unit entitles any number of users to run the specified Software upon output created by one single Oracle Instance. An Oracle Instance is defined as one set of Oracle background processes/threads and a shared memory area, which is memory that is shared across those processes/threads running on a single computer. For example, a 3-node Oracle Real Applications Cluster (RAC) system, is a 3-Instance system.

Maintenance License. An active Maintenance License entitles a licensee to download and install, at no additional fee, all new versions of the specified Software (for the licensed platforms) that are released within eleven (11) months from the date of purchase. To download and install a version released after this period requires an additional fee.

Oracle is a registered trademark of Oracle Corporation.