|
|
JL Computer Consultancy |
|
Single-day Tutorials |
Finding the time for three consecutive days of seminar-based training is not easy, and you may not be interested in all aspects of making Oracle work better for you - especially if you have a legacy system to support. To help you out, I have prepared several separate, stand-alone, training courses focusing on particular aspects of Oracle performance.
The courses are lecture-based, and consist of four sessions (9 in the case of the two-day event) of about 80 – 90 minutes each. The materials for the course consists of printouts of the Powerpoint slides, roughly 40 per session, with a footnote on each slide highlight particular points made by the slide.
Dates
|
Dates currently arranged for the next |
|
|||||||
|
|
|
|
|
|
|
|
|
|
|
Dates currently arranged for International Tutorials |
|
|||||||
|
|
|
8th / 9th Feb |
|
|
( |
|
||
|
|
|
8th / 9th Mar |
|
|
( |
|
||
|
|
|
Late March |
|
|
( |
|
||
|
|
|
8th / 9th Apr |
|
t.b.c |
|
( |
|
|
|
|
|
April 21st |
|
|
( |
|
||
|
|
|
April 27th/28th |
|
|
( |
|
||
|
|
|
17th / 18th May |
|
|
(Trutek) |
|
||
|
|
|
19th May |
|
|
NYOUG & John Goodhue |
|
||
|
|
|
20th May |
|
|
NEOOUG & John Goodhue |
|
||
If you would be interested in attending any of the courses, please send an email to jonathan@jlcomp.demon.co.uk. Please note, any email will be taken purely as a notification of interest, not as any sort of commitment.
|
Is there a go_faster parameter ? Which parameters affect the Cost Based Optimizer (CBO) most? Just how does Oracle decide whether or not to use an index? Why does Oracle do a Hash Join instead of a Nested Loop? Why is Oracle ignoring my hints? All these, and many other questions about the working of the Cost Based Optimizer will be answered by this tutorial. |
|
|
One of the most important, yet overlooked, aspects of efficiency is the need for a good indexing strategy. Indexes improve the speed of queries, but increase the cost of DML. Given a thorough understanding of indexes you have the option for deciding the optimal set of indexes to suit the needs of your application. Everything you might want to know about indexes will be in this tutorial. |
|
|
The users are complaining right now ! How do I find the problem - what's the short term fix - what's the long term fix? The overnight batch ran on until 9:30 this morning when it's supposed to complete at 5:00 a.m. - what went wrong? Good performance starts with good design - which means you have to know how much data you have, where to put it, and how you're going to use it. Working out the best strategies is called tuning. After go-live you don't do tuning, you do trouble-shooting, and the strategies have to change to suit the circumstances. In this one-day course, we learn how to think around design issues (the tuning) as well as looking at strategies - and some fixes - for addressing implementation problems (the trouble-shooting). The final session of the course examines some of the dynamic performance views, what you can do with them, and how their contents are reported in the Statspack and AWR reports. |
|
|
This is a course aimed at developers who wants to learn what the database can do to help them produce better applications. Its focus is very much on the database engine - what the engine hopes to be given as a workload. The tutorial covers overheads, structural features, coding strategies, indexing, and execution plans. Although the course is intended for the more senior developers, it can also act as a rapid introduction to critical mechanisms of the database that the junior DBA should need to know.. Note – the material is about the database and using it properly; it is not about programming and programming languages. This day is most suitable for the enthusiastic developer who wants to be more aware of how the database and application can co-operate to work as efficiently as possible. |
|
|
One of the critical skills for tuning SQL lies in understanding what is happening, and what is possible. Oracle gives you several tools for examining (or predicting) the execution path for a query. In this tutorial we look at the available options, examine their strengths and weaknesses, and learn how easy it is to understand what an execution plan is telling you.. |
|
|
This tutorial has two targets. First: how do you improve the performance of a production system by attacking inefficient SQL; secondly: how do you design a system so that you can get data into and out of it efficiently. For most people, the design has already happened, nevertheless you need to be aware that the two targets overlap significantly. As you struggle to address a particularly inefficient piece of SQL you need to be aware that there are structural features of the database that may help you find a cost effective solution despite the problems imposed by the SQL itself. The course will focus on the SQL – falling back to structural issues as an aid to improving the mechanical efficiency of the SQL solution – and will spend some time explaining the pros and cons of various structural options as they become relevant to the SQL. The course will cover methods of reviewing data distribution patterns, use of indexes, use of views (stored and inline), analytic functions, subquery factoring, as well as a brief review on hints and execution plans. |
|
|
This two-day course is a combination of sessions from the CBO, Explain Plan, and Indexing tutorials. The aim of the course is to give you the information you need to understand the principal mechanisms that drives the optimizer's choice of execution plan. We examine the basic arithmetic, the metadata that the optimizer needs if it is to understand how to handle your data, and the way you can structure your data and code to help the optimizer. |
Format
The tutorials will be lecture style with Powerpoint presentations, using a classroom-style arrangement. Each tutorial will consist of four presentations of about 1.5 hours each (with a short break in the middle of each, if someone reminds me to take it). The Powerpoint slides which will be available to the delegates in printed form at the start of the tutorial. There will be breaks for refreshment and lunch between presentations.
I can also run any of the courses as an in-house tutorial for larger companies who have suitable facilities on-site.
Arrangements for international events will generally be similar - but where the event is organised by a third party you may expect some variation in the format.
|
Understanding and Assisting the
Cost Based Optimizer – Agenda |
|
|
Session 1 1.5 hours |
Why isn't Oracle using my index? In this session we aim to acquire a visual, rather than highly mathematical, understanding of when Oracle will ignore an index. We start with a discussion of the strategic direction that we should be going with the cost based optimizer and then examine a simple example to pin-point the traditional reasons why the cost based optimizer in versions prior to 9.0 could so easily produce inappropriate execution plans. We examine a couple of commonly used tuning parameters which give Oracle better information about our system and allow it to identify the appropriate execution path more frequently, identifying the risks of using these parameters too extravagantly, and end with a demonstration of how using CPU costing solves the traditional problems with much less risk of error. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
Mechanisms of Joins We start by asking why it is so important to examine and understand join mechanisms and examine some of the transformations that convert complex queries to simple join forms. Then we look at details (and anomalies) of the nested loop, sort-merge, and hash joins; identifying strengths, weaknesses, and costs of each in turn. In the case of sort-merge and hash joins, we examine trace events, and their output, that allow us to investigate what is happening when response times become extreme. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Selectivity, Joins, and Hints In this session we move on from the simple example of the first session to investigate what happens in the more general cases of using an index. We extend the arithmetic to the calculations involved in joining tables, and note some of the anomalies and problems that joins can cause. Finally we ask the question “What is a hint?” and try to answer the question by examining what really happens inside the optimizer when we start adding hints to our SQL. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
Maximising the Truth If there is any information we can give the optimizer about our data, we should do so; otherwise it will be less able to produce the appropriate execution. We start this session by showing how a little extra information can help the optimizer find new execution paths. Then we examine the optimizer’s dependence on statistics, and its need to get a correct numeric representation of your data. We move on to see how strange data patterns, and bad database design, can stop the optimizer from choosing a sensible execution path, and end with a couple of techniques for encouraging the optimizer to do what we want. |
|
|
|
|
Session 1 1.5 hours |
The Structure of Indexes What is a B-tree index, what is a Bitmap index and why are the two so different? How does Oracle build indexes? Are there differences in reverse indexes, function-based indexes, cluster indexes, global partitioned indexes, indexed organized tables, and secondary indexes? With the aid of a few block dumps and carefully monitored test results we examine the way that different types of indexes are built and behave. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
The Uses of Indexes How many different ways can Oracle take advantage of an index. We have primary keys and unique keys - do we need unique indexes to enforce them. We may have problems with foreign keys if we don't have related indexes. We need optimal access paths - but will indexing help and what will it cost? When will Oracle use an index to accelerate a query? Can we assist performance by understanding the mechanisms. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Popular misconceptions There are several surprising ideas floating around the Internet about how Oracle indexes work. Sometimes these ideas are irrelevant and have no use or impact. Sometimes they have a serious impact on your effectiveness. They may cause you to do work that need not, or should not, be done. They may introduce down-time that you do not need. They may make you avoid a feature that is really the perfect feature for your application. This session examines some of the more popular and potentially damaging misconceptions about indexes. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
Maximum benefit, minimum cost We all know that indexes are supposed to make things go faster - but sometimes we forget that they have maintenance costs - of two types. How do you identify the optimal set of indexes for your application? How do you identify redundant indexes? How do you check whether or not an index has become inefficient and needs to be rebuilt - and how do you rebuild it most cost-effectively? This session discusses strategies for ensuring that you don't waste effort supporting indexes that shouldn't exist, and don't waste valuable batch time rebuilding indexes that don't need it. |
|
|
|
|
Note: There is an
alternative version of this day covering sessions one, two and four, and
closes with a session where the attendees are invited to supply production
reports from Statspack or the AWR for public
analysis and review. There reports
should be the text option generated by spreport.sql
or awrrpt.sql, ideally based on a snapshot interval
of one hour, running (for statspack) at level
seven. |
|
|
Session 1 1.5 hours |
Trouble-shooting or Tuning What's the difference? What are the strategies. Why tuning is hard, trouble-shooting is easy and design is important. Key targets for producing a well-tuned system on day one. How to approach design, and plan for trouble-shooting, and how to be on the alert for activities where the resource utilisation is unreasonably high. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
Looking for Problems If you know the sorts of problems your design is going to face, then you There are three major classes of problem, but just one source of (Oracle) information. It’s important to recognize what class of problem you’re attacking and pick the best way of using the available information to identify the source of your problem. We outline examples of the three classes and the corresponding approaches, and end with a detailed example of problem-solving. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Tactical Fixes Methods, workarounds, dirty tricks and parameters for dealing with classic performance problems when the system is in production. Some of the examples we cover are specific to particular production issues, some are generic quick fixes that can be applied across the board. All fixes need careful examination of costs, risk, and benefits. In this session we consider examples of strategies that are most likely to be worthwhile. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
V$, X$ and Statspack/AWR It's a good idea to be familiar with just a few of the dynamic performance views - and there are a couple of items in the still hidden away in the X$ objects that can add be useful in special cases. This session will present some ideas on how to take advantage of the dynamic performance objects directly, and then move on to some of the more useful information they supply. In particular, we review the Statspack and AWR interfaces to the dynamic performance views - closing with a case study of using Statspack to solve a performance problem. |
|
|
|
|
Session 1 1.5 hours |
Overheads A basic introduction to the costs of the infrastructure that makes Oracle what it is. Topics include connection costs, network traffic, parsing and optimising costs, undo generation and use, and redo. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
Coding Strategies There are good ways to code and bad ways to code. This session demonstrates the problems with the bad coding strategies and shows how much difference it makes when you do it right. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Indexing A description of indexes, largely devoted to B-tree indexes, talking about structure, how they work, and the paths Oracle is able to take through them. Indications of how to get the most out of indexes, and minimise the cost of indexing, with warnings about side effects of the different strategies for making an index more useful. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
Execution paths In this session, we take an introductory view on how the cost based optimizer works, including a brief overview of the different join mechanisms. Then we examine execution paths and the developer-oriented ways of acquiring them. Finally we learn how to read and interpret execution paths. |
|
|
|
|
Session 1 1.5 hours |
How to find execution paths. There are several different methods for generating, or examining, execution paths, from the simple autotrace in SQL*Plus through to interrogating the dynamic performance views such as v$sql_plan_statistics_all in great detail. The different methods have their own strengths and weaknesses, and these vary with version of Oracle. In this first session, we discuss the different options, what’s visible in them, and to choose the most appropriate one for the task in hand. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
Intepreting Plans Once you’ve got an execution plan, how do you decide what it means ? An often-quoted clue is that ‘it starts near the top and over to the right’. This hint is a consequence of a simple algorithm for reading execution plans – no matter how complex – in a fashion that can be described in just two steps. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Problems with Plans Although there are several ways of acquiring execution plans, none of them is perfect. Some options simply omit critical details; some options are unable of guaranteeing the truth. In this session, we look at the various omissions and errors that can make you jump to the wrong conclusion when you are trying to understand how a query might run. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
Advanced Features Once you have mastered the standard features of execution plans, you can easily extend your understanding to any new feature that Oracle throws at you. In this session, we take a look some of the features that are not seen so commonly in production systems, namely , distributed joins, parallel execution, and partitioned tables. |
|
|
|
|
Session 1 1.5 hours |
Background. In which we examine the need for a global view point and a general strategy for minimising work while recognizing that there will always be a conflict between local and global optimization. A quick review of use of indexes, including some of the less common uses and common errors in index design. Closing with the need to be able to investigate and understand the data and business requirements. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 2 1.5 hours |
Basic Practices How to approach the task of translating business requirements into SQL. The need for clarity in code, with a suggested set of standards for presentation to improve ease of understanding. An introduction to the essential strategy for writing SQL that gets its result by doing the minimum work, and a model for investigating and re-engineering badly performing statements. |
|
Break – Lunch and informal discussion: 1 hour |
|
|
Session 3 1.5 hours |
Sundry Methods A brief discussion of the mechanics (and costing) of nested loop and hash joins, and why it’s so easy it is for the optimizer to pick the wrong option. A few examples demonstrating classic problems, and solutions that take advantage of some of the slightly exotic options of SQL and Oracle’s available features. And a few closing comments about the need for the front-end code to co-operate with the database engine in certain circumstances. |
|
Break – coffee and informal discussion: 30 minutes |
|
|
Session 4 1.5 hours |
More Methods Fixing the code without touching the code – the latest options, including a little time looking at the ways in which you can use OEM to assist with tuning. Problems and options with knowing internal mechanisms and taking advantage of them. A look at a couple of popular requirements, and strategies for addressing them as efficiently as possible. And a final look at the threats of hints and how to use them properly. |
|
|
|