Reverse Engineer Machine Learning To Develop Performance Monitoring Rules - This presentation demonstrates how to create supervised machine learning model to recognize patterns of poor performance, but then extract the rules in plain English. All this is based on sysmetric data and Python within a Jupyter Notebook.

Let's Build An Automated Anomalous DB Activity Detector Using Machine Learning - This presentation demonstrates how to build an anomaly detection model based on Oracle AWR symmetric data using Python within a Jupyter Notebook.

How To Prove The Problem Is Not The Database Using ASH Data - The combination of enabling ASH to temporarily collect "non-active" sessions, and step-by-step watching of an Oracle session while paying close attention to the "SQL*Net message from client" event, we can prove a performance issue resides outside the database!

Powerful: Practical Machine Learning In The Hands Of The Oracle DBA/Developer - In this presentation I'm going to introduce you to the world of ML from a DBA perspective. This includes understanding what ML is, why we use it and why now. I will lead you through two use cases, each using a different ML algorithm. The first at a high level and the second at a detailed level.

How To Analyze SQL Run Times By Plan Using ASH Data - In this presentation, using ASH data, I will show you how to manually infer SQL run times. Then I'll show you how to use a simple yet flexible SQL script to analyze ASH data, infer SQL run times and report the results... even at the execution plan level. Next I'll show you how to analyze the run time samples using the free statistical package R.

Top 10 Ways To Totally Mess Up An AWR Analysis - There are some obvious ways to totally invalidate your analysis. But, there are also some subtle mess-ups that fool you into thinking your time-based analysis is rock solid...yet it's worthless and probably misleading. Join me for a light-hearted session as we explore what NOT to do.

How To Build A Performance Indicator Using Machine Learning - Learn how to build a supervised learning classification machine learning model with Python within a Jupyter Notebook, including how to train the model to recognize and "understand" a virtually unlimited number of performance and business related characteristics.

How To Do Impossible "Log File Sync" Performance Analysis With ASH Data - In this presentation I will show you how to find the DML SQL associated with a commit, using ASH data.

How To Use An AWR/Statspack Report To Tune Oracle Systems - In this presentation I will teach you how, step-by-step, to take an AWR report or a Statspack Report, extract only the information you need, build a diagnostic framework to expose the biggest tuning opportunities and then develop solutions that directly address them.

Creating Presentations People Love And You Enjoy Giving - Over the years, I have developed a kind of method or process I go through that has worked well for me. In this presentation, I will share this process with you.

How To Diagnose Random Oracle Performance Incidents Using ASH Data - In this presentation, I begin with why ASH (active session history) is so powerful, how it works and step-by-step how to use ASH to solve complex performance incidents. Using only four free scripts, I will lead you through the incident analysis process, including how to visualize the incident using the free statistics package R.

Parsing With Plan Change Detection And Analysis Using ASH Data - In this presentation I will show you how to use ASH data to pinpoint SQL plan changes and how to detect parsing issues before they show up in the wait interface.

Understanding Oracle Memory Serialization Control (Latches and Mutexes) - To get the upper hand on serialization control, in this presentation I will explore the lock (a little), the latch (a lot), and the mutex (a whole lot). Topics include performance diagnosis, how Oracle implements latches and mutexes, and the related internal algorithms. Special attention is given to the library cache mutex operations.

How To Choose Your Best Oracle Performance Analysis Strategy - So, what is your best analysis strategy? As Oracle DBAs, we have two fundamental strategies: time-based and active session sample based (ASH). Knowing their differences and how to exploit them gives you the optimization edge.

Laugh Your Way To Understanding Oracle, Queuing Theory & Performance - You can laugh your way to understanding Oracle, queuing theory and performance analysis. With only the basics, we can use queuing theory to derive targeted performance solutions and goals, filter and evaluate any performance solution thrown at us and help non-technical people understand why our solutions make perfect sense.

Maximizing Oracle Buffer Cache Throughput - Part of our job as an Oracle DBA is to understand the buffer cache and maximize the work flowing through it. This is a very practical yet deep internals presentation, filled with amazing discoveries about how Oracle works.

Navigating Your DBA Career In The Oracle Cloud - While some things never change, some things do. And it's the things that change that can have a huge impact on us and our families. This presentation will help you understand what has happened, what is happening, what you can expect and what you can do today to ensure you are positioned to thrive in a cloud world.

How To Do An Oracle Time-Based (OTBA) Analysis - Unless you have a quantitative approach, your analysis will always be subjective. In this session you will learn how to set up and perform a time-based analysis based on an AWR or Statspack report.

How To Tell If There Is A Memory Shortage And What To Do About It - An OS memory shortage can cause bizarre and intense performance situations. In this presentation we will explore different ways to tell if there is a memory shortage and how this affects Oracle performance.

Understanding And Resolving Oracle Cache Buffer Chain Wait/Latch Contention - When a server process needs to access a buffer, it must first access the cache buffer chain (CBC) structure to determine if the buffer is in the buffer cache. While serialization control protects the CBC structure, given high concurrency and just the right workload mix, performance can be a problem. In this presentation you'll learn the relevant Oracle internals, quantitative methodical problem diagnosis, and solutions.

How To Start Blogging To Propel Your Career - Craig discovered a kind of formula or process to help DBAs get started with blogging in a way that meets their objectives. It is not as difficult as you might think, but if you just "wing it" you will likely become frustrated and disappointed with the results.

Let's Build Something! Building Your Own ASH Toolkit - In this session I'll show you how to build four highly flexible ASH-based scripts; summary, top wait, top CPU and session time-line. I'll even demo the scripts I just created on a real Oracle workload!

Creative Oracle 12c Redo Maneuvers - There are a number of creative maneuvers Oracle DBAs can implement that increase the redo flow, reducing log file sync and log file parallel write wait times. This presentation explores such maneuvers, as well as redo process and memory architecture.

Merging Oracle Time-Based Analysis With Queuing Theory For Practical Results - What is unique about this presentation is Craig combines Oracle Time-Based Analysis with queuing theory. In doing so, using MS Excel, Craig (and the students) create response time graphs from production AWR reports AND anticipate a SQL statement's elapsed time by tuning it!

Identifying High Risk SQL Before The Upgrade - The differences between the two systems can cause run time differences. The trick is creating an apples-to-apples QAT-to-Production comparison. This session will focus on how to factor in and remove the problem of CPU speed differences and workload intensity differences.

Cursors, Child Cursors And More Child Cursors - Learn about Oracle library cache internals focusing on cursors and child cursors. Multiple child cursors are essential for many applications, but when hundreds or thousands are created, massive performance problems can occur. Craig covers how multiple child cursors occur, how to detect if it's a problem and what to do about it.

How To Understand Oracle CPU Consumption: The Missing Link - Oracle has the world's most advanced database wait time instrumentation. However, Oracle does not instrument CPU consumption. In this presentation, you will learn how to determine where Oracle is consuming CPU and how to use this in formulating targeted performance solutions.

The Average Deception - The word "average" immediately conjures up pictures of a nice, evenly balanced figure with the results being higher sometimes and lower sometimes. However, in Oracle performance analysis, this is rarely the case. This reality mismatch, observing it and using it to our analysis advantage, is what this presentation is all about.

Unit Of Work Time-Based Analysis - Most Oracle performance analysis is time-based with the focus on "total time." For example, how much time does it take to process a SQL statement. But there is a deeper way to understand performance. By focusing on the time to process a single piece of data work (think: LIO), we open up an entirely new arena for advanced performance analysis.

Two Applications Of Oracle Serialization Control - Let's apply Oracle memory structure access control in non-traditional areas. First, Craig will focus on cache buffer chain performance analysis. Second, Craig will teach you why tuning serialization control broadly impacts performance and how you can use this knowledge to anticipate the elapsed time of a specific SQL statement.

Stop The Guessing! Using Time-Based Analysis To Improve Oracle Performance - Unless you have a quantitative approach, your analysis will always be subjective, and may even miss the root cause of performance issues. In this presentation you will learn how to set up and perform a time-based analysis, where to get performance data, how to construct your diagnosis framework and how to derive targeted performance solutions.

Calculating OS CPU Utilization From V$ Views, AWR and Statspack - Oracle collects the data we need to calculate the OS CPU utilization. The data is stored in v$ views and both AWR and Statspack tables. This short presentation will introduce the utilization concept and then demonstrate how you can quickly calculate or even “eyeball” utilization from an AWR or Statspack report using two different methods.

Total Performance Management - This is a timeless paper. It is as relevant today as it was way back in 1995... As management expectations of the DBA increase, the need to maintain optimal performance, with every increasing application complexity and increasing data volume, has placed many DBAs in a frenzy.  This paper explains how to get out of this mess.