36

V8-V9 design for high performance workshop

Date: 
Thursday, 26 June, 2008 (All day)
Speaker: 
Jan Henderyckx

Objectives:
Understanding which DB2 features and techniques exist to get the best performance and availability out of your application and system. Know which knobs exist at the database design level to get the most of your system. Be equiped with application design patterns that can deal with the most challenging application requirements.

Outline:
DATABASE DESIGN

REORDERED ROW FORMAT
New Datatypes
APPEND table attribute
NOT LOGGED TABLESPACES
IMPLICITLY HIDDEN COLUMN
INDEX ON EXPRESSION
LARGER INDEX PAGE SIZE
ASYMMETRIC INDEX PAGE SPLIT
RANDOM index keys
XML Data Type

PROGRAMMING TEMPLATES
Techniques to generate numbers
Native Stored Procedures
Instead of Triggger
Multi-row
Browsing
MERGE STATEMENT
Truncate

MANAGING CONCURRENCY
Overview of relevant zparms
OPTIMISTIC LOCKING SUPPORT
SKIP LOCKED DATA ON SQL statement

DESIGNING FOR 24x7
Clone Tables
Beyond Clone Tables

Full Abstract:
During this seminar Jan will take you through the important aspects that influence your application performance and high availability. In the last two releases many new features have been added that will have an important impact on the efficiency of your applications. There is a lot of information available in books and articles but this ODS is based on real life cases, many which have been implemented by Jan in customer engagements. All of this with the fundamental principle that we don’t tune and design for the sake of the tuning but to make sure we meet the application objectives.
 

Application performance & design 2008

Date: 
Thursday, 26 June, 2008 (All day)
Speaker: 
Klaas Brant

Objectives:
Get more knowledge about how DB2 processes applications. With this knowledge tune applications for maximum performance.

Outline:
Part 1 DB2 Accounting trace

DB2 trace: classes & destinations
What is (not) in the DB2 accounting record
Nested activity (Stored Procedures, UDFs & Triggers)
Your Performance Indicators
CPU
Getpages
Wait time (not lock)
Lock time
IO's
Sort

Part 2 SQL Access Path
Explain: What does this (not) tell you
DSNSTMT table, use it!
Can the statement be improved?
Tuning toolbox
Groundrules for tuning
SQL Tuning
Forcing DB2 (optimize for / Hints)
Optimizer rewrite, when and how
Asking DB2 to reconsider
More and better stats
Bad thing you should not consider
Parallelism, why and why not!
How about dynamic SQL?

Part 3 Hints and tips for a better performance
Better SQL
Beter Database design
Improving your application logic
Batch Issues

Full Abstract:
There are many, many things that influence your performance. Where to find your performance indicators? How to interpret them?
This seminar looks at collecting and interpreting performance figures, looks at alternatives and discusses what should be done in order to obtain better performance. The last few releases of DB2 gave us some great things (e.g. multi row insert and fetch, and volatile tables). We will review these things in order to improve our applications. You will go home with some new idea's to improve your application.

V8/V9 SQL/Database Challenges and Solutions

Date: 
Thursday, 26 June, 2008 (All day)
Speaker: 
Susan Lawson

Objectives:
Discuss the importance of smart SQL and applications
Discuss V8 SQL and application features and show examples
Discuss V8 features for database performance
Discuss challenges still remaining in V8 SQL/applications and databases
Look at some new V9 SQL/application and database features and their benefits

Outline:
V8 SQL/Application Performance Features
Multi-Row Fetch
Recursion
Common Table Expressions
INSERT within a SELECT
Scalar Fullselect
Distribution Statistics
XML Scalar Functions
Multiple Distinct
Matching Predicates
Parallelism and Sorting
V8 Database Performance Features
Volatile Tables
Data Partitioned Secondary Indexes
V9 SQL/Application Features
UPDATE/DELETE with a SELECT
Optimistic Locking
Index on Expression
Histogram Statistics
Order By and Fetch First in Subselect
Instead of Triggers
INTERSECT and EXCEPT
Skip Locked Data
Caseless Expressions
Native SQL Procedures
XML Support
Star Joins Performance
New Functions
V9 Database Performance Features
Better Index Page Splitting
Index Compression and Large Pages
DPSI Enhancements
Clone Tables
Append on Insert
Truncate Table
No Log Tables

Full Abstract:
In this seminar we will look at some of the performance features in V8 and how they have been successfully used in our applications and databases to meet some of our current performance challenges. Those challenges, of course, being performance and availability. We will discuss some of the things we still desire in terms of application and database performance and take a look at V9 to see what features are available to further improve our performance and availability of our applications. We will look at how we may take advantage of those features.
 

DB2 Version 9 for z/OS In Depth

Date: 
Thursday, 26 June, 2008 (All day)
Speaker: 
Phil Grainger

Objectives:
This seminar will give you a good all round understanding of the new features and functions of DB2 9 for z/OS

Outline:
New Objects in DB2 9
DB2 9 performance improvements
New security options
Support for XML data storage
.. and more!

Full Abstract:
This one day class will look through all of the new items IBM have released in DB2 9 for z/OS. From new relational objects (such as table clones, universal tablespaces and trusted contexts) to the new non-relational XML document storage, Phil takes us though the whold range.
Even if DB2 9 is still some way in your future, there is almost certainly something here of interest.

How to Choose A DB2 Tool Solution for YOU?

Date: 
Wednesday, 25 June, 2008 - 13:15 - 14:15
Speaker: 
Phil Grainger

Objectives:
To gain an understanding of how to choose a software tool to assist with your daily jobs. This presentation is not aimed at any particular tool choice, nor at any particular vendor (honest!) but is a summary of my experiences over the years of why people choose some tools over others

Outline:
Who makes the choices anyway?
What makes a good choice?
What is “value for money”?
Why do some vendors “muddy the waters”?
Who has the most negotiating power?

Full Abstract:
You might think that finding the right software tool to help you do your job would be simplicity itself. This presentation looks at some of the ways you may be losing control of your choice without even realising it and how to get that control back. The session will be a completely open and independent consideration of the processes you should be working through as you make your tools choice

Top 10 Ways to Waste CPU

Date: 
Wednesday, 25 June, 2008 - 12:15 - 13:15
Speaker: 
Susan Lawson

Objectives:
Discuss the top 10 ways CPU is wasted
Discuss the reason why this is the case
Discuss the technical aspects of DB2 performance
Determine how to find these types of problems
Show and discuss how to prevent these problems

Outline:
1.Overuse of DB2 Compression
2.Excessive Locking
3.Processing of Data in an SQL Statement
4.Overuse of DISTINCT
5.Inefficient Predicates
6.Inappropriate BIND Parameters
7.Blind Updates
8.Random Distribution/Access
9.Too Many Indexes
10.Black Box I/O Modules

Full Abstract:
During the past several years we have worked with several companies to design high performance databases and applications and have done extensive testing and research on ways to minimize I/O and CPU. Many of those applications and databases are faster than we ever could imagine. However, those are the exception. Too often we see shops with critically CPU constrained environments. Most of these shops have one thing in common: They are wasting CPU. There are some very common ways to do this and we need to look at how you can safely and effectively reduce CPU usage in a DB2 z/OS environment. This presentation will look at the top 10 ways we have seen customers waste enormous amounts of CPU unnecessarily (and unknowingly). A few tuning efforts here and there can improve overall performance and reduce CPU consumption.
Of course every environment is different and processing needs will vary. However, there are still ways to reduce CPU usage through some very practical performance tuning efforts and better design efforts early in the development.
 

Bringing your DB2 to V9 (and V8)

Date: 
Wednesday, 25 June, 2008 - 11:15 - 12:15
Speaker: 
Jan Henderyckx

Objectives:

Outline:

Full Abstract:
Many DB2 systems are still running as if it where the 90's. There are many reasons for this lack of adaptation but one of the most important ones is lack of time. This means in many cases not lack of time but simply lack of priority. This presentation helps you to look at problem and to build a business case around it. How much will you gain if you make this change. Putting the numbers behind the change will greatly help you with the sell.

Making the database group survive ITIL, Service management, etc

Date: 
Wednesday, 25 June, 2008 - 10:00 - 11:00
Speaker: 
Jan Henderyckx

Objectives:

Outline:

Full Abstract:
Do you also find yourself being defined in terms of services? How can you best deal with that and structure your work to fit in such an organisation. Jan will give practical examples how you can define services for your DBA team and how you can more effectively manage your DB2 objects by defining proper responsibilities. The presentation includes a sample service registry for a DBA team and an example risk and gap analysis.

Universal Table Spaces, what are they?

Date: 
Wednesday, 25 June, 2008 - 09:00 - 10:00
Speaker: 
Klaas Brant

Objectives:
Understand the new Universal Table Space and underatnd IBM intentions for the future

Outline:
What we have now and what’s wrong with it
The new type of table space: universal
Partition by Growth
Partition by Range
Interaction with utilities

Full Abstract:
DB2 9 comes with a brand new table space type called Universal table space. Why do we need this tablespace? What is the diference between Partion by Growth and Partion by Range type. What are IBM plans for the future regarding table spaces. This session explains all about these new features.

Choosing the Right Indexes

Date: 
Tuesday, 24 June, 2008 - 16:15 - 17:00
Speaker: 
Zeljen Stanic

Objectives:

Outline:

Full Abstract:
Creating indexes on DB2 tables is one of the simplest ways to enhance performance.
It is also one of the easiest things to do - you just type CREATE INDEX right?
The skill is in knowing WHICH indexes to create - after all, too many indexes can degrade application performance.
Unicenter Index Expert from CA makes it easy to determine WHICH indexes are RIGHT for your table.
By taking into account statement execution frequency, a set of indexes that are optimized for your workload are recommended.
Index Expert will determine the ideal set of indexes for any given DB2 table
A number of events could trigger a need to consider or reconsider the current indexing strategy
The development or purchase of a new application
The discovery of a performance issue (such as might be highlighted by CA Plan Analyzer, CA Detector or CA Insight for DB2)
A need for a ‘health check’ of an existing application to ensure that the assumptions made during the database design are still valid 

Syndicate content  

sqladria

SQL Adria is the independent, non-profit organization that gathers relational database users for Croatia and Slovenia. It was founded 1994. and in the same year it has become the regional user group.

Follow us

facebook
linkid
twitter

 

Events in 2024

VODICE -  10th - 13th June 2024