Oracle Database 12 c : Real Application Testing Overview

Oracle Database 12 c : Real Application Testing Overview
Oracle Database 12c: Real Application Testing Overview
An Oracle White Paper
October 2013
Oracle Database 12c:
Real Application Testing Overview
Oracle Database 12c: Real Application Testing Overview
Introduction ....................................................................................... 3
Real Application Testing .................................................................... 3
SQL Performance Analyzer ........................................................... 4
Database Replay ........................................................................... 7
Conclusion ...................................................................................... 12
Oracle Database 12c: Real Application Testing Overview
Introduction
The Oracle database is the market-leader and the preferred database for hundreds of thousands
of enterprises as well as application developers and database administrators worldwide. Over the
years, enterprises have come to rely on the Oracle database to provide unparalleled performance
and reliability. Oracle continues to raise the bar with Oracle Database 12c with extensive support
for consolidation. Designed for data center environments that are rapidly evolving and changing
to keep up with the demands of the business, Oracle Database 12c allows businesses to adopt
new technologies quickly while minimizing risk.
Real Application Testing
Today, enterprises have to make sizeable investments in hardware and software to roll out
infrastructure changes. For example, a data center may have an initiative to move databases to a
low cost computing platform, such as Oracle Enterprise Linux. This would, traditionally, require
the enterprise to invest in duplicate hardware for the entire application stack, including web
server, application server and database, to test their production applications. Organizations
therefore find it very expensive to evaluate and implement changes to their data center
infrastructure. Additionally, in spite of the extensive testing performed by customers, unexpected
problems are frequently encountered when a change is finally made in the production system.
This is because test workloads are typically simulated and are not accurate or complete
representations of true production workloads. Data center managers are therefore reluctant to
adopt new technologies and adapt their businesses to the rapidly changing competitive pressures.
Oracle Database 12c Real Application Testing option addresses these issues head-on with two
complementary solutions, SQL Performance Analyzer and Database Replay.
3
Oracle Database 12c: Real Application Testing Overview
SQL Performance Analyzer
Changes that affect SQL execution plans can severely impact application performance and
availability. As a result, DBAs spend enormous amounts of time identifying and fixing SQL
statements that have regressed due to the system changes. SQL Performance Analyzer (SPA) can
predict and prevent SQL execution performance problems caused by environment changes.
SQL Performance Analyzer provides a granular view of the impact of environment changes on
SQL execution plans and statistics by running the SQL statements serially before and after the
changes. The SQL Performance Analyzer report that is generated outlines the net benefit on the
workload due to the system change as well as the set of regressed SQL statements. For regressed
SQL statements, appropriate execution plan details along with recommendations to tune them
are provided.
SQL Performance Analyzer is well integrated with existing SQL Tuning Set (STS), SQL Tuning
Advisor and SQL Plan Management functionality. SQL Performance Analyzer completely
automates and simplifies the manual and time-consuming process of assessing the impact of
change on extremely large SQL workloads (thousands of SQL statements). DBAs can use SQL
Tuning Advisor to fix the regressed SQL statements in test environments and generate new
plans. These plans are then seeded to SQL Plan Management baselines and exported back into
production. Thus, using SQL Performance Analyzer, businesses can validate with a high degree
of confidence that a system change to a production environment in fact results in net positive
improvement at a significantly lower cost.
Examples of common system changes for which you can use the SQL Performance Analyzer
include:

Database upgrade, patches, initialization parameter changes

Configuration changes to the operating system, hardware, or database

Schema changes such as adding new indexes, partitioning or materialized views

Gathering optimizer statistics.

SQL tuning actions, for example, creating SQL profiles

Database consolidation testing using Oracle Multitenant Databases or schema consolidation
methods
SQL Performance Analyzer functionality is available through DBMS_SQLPA API and through
Oracle Enterprise Manager interface. There are various custom workflows available through
Enterprise Manager and the Guided Workflow encompasses all of them and provides complete
control and flexibility at every step of the process. Using the SQL Performance Analyzer Guided
Workflow involves the following 5 main steps:
1.
Capture the SQL workload that you want to analyze with SPA. The Oracle database offers
ways to capture SQL workload from several sources, such as cursor cache and Automatic
4
Oracle Database 12c: Real Application Testing Overview
Workload Repository, into a SQL Tuning Set (STS). This would typically be done on a
production system and the STS would then be transported to the test system where SPA
analysis will take place.
2.
Measure the performance of the workload before a change by executing SPA on the STS.
Very short running queries are executed multiple times and their statistics are averaged to
eliminate variations due to buffer cache state and other noise factors
3.
Make the change, such as database upgrade or optimizer statistics refresh.
4.
Measure performance of the workload after the change by executing SPA on the STS again,
as in step 2.
5.
Compare performance of the two executions of the SQL tuning set to identify the SQL
statements that have regressed, improved, or were unchanged.
Figure 1. SQL Performance Analyzer report
The SPA comparison report in Figure 1 shows significant performance improvement of overall
SQL workload after the proposed system change but with a few execution plan regressions. SQL
Performance Analyzer takes into account the number of executions of a SQL statement when
measuring its impact. A SQL statement that completes in seconds but is frequently executed may
have a higher impact on the system than a long running statement executed only once. SPA takes
these factors into account when predicting overall performance improvements and regressions. If
any regressions are encountered, SPA allows the user to fix them using SQL Tuning Advisor or
with SQL Plan Baselines, a plan stability feature introduced in Oracle Database 11g.
5
Oracle Database 12c: Real Application Testing Overview
SPA Quick Check is an enhancement to SQL Performance Analyzer introduced in Enterprise
Manager Cloud Control 12c Release 3, Plug-in Release 1 and can be used on Oracle database
versions 11.2 and above.
SPA Quick Check enables verification of routine DBA operations that affect database
performance on production databases with a single click of a button. SPA Quick Check runs
SQL Performance Analyzer in an optimized mode that is suitable for running on production
systems. In this mode, it uses minimal system resources by first identifying SQL with plan
changes and then test-executing only those SQL, thereby consuming only a fraction of system
resources. SPA Quick Check further limits unusual resource usage by placing run-time limits for
each SQL and uses Resource Manager to ensure that only a limited amount of system resources
are utilized by this task. All these optimizations make it possible to run SPA on production
databases for routine DBA tasks such as optimizer statistics gathering, init.ora parameter
changes, etc.
SPA Quick Check enables DBAs to run quick validations for any system change they are
required to make to a database and helps improve quality of service of their mission critical
databases.
SPA supports numerous other features that help assess system changes, these are briefly
described below:
1.
SPA helps estimate the IO reduction that can be accomplished by migrating to Exadata
server but without actually requiring you provision the hardware. This can be used to
identify potential workloads/systems that are good candidates for Exadata migration.
2.
SPA supports comparing the performance of two similar workloads or STSs – this
functionality is useful when you have mechanisms such as load testing scripts or Oracle
Application Testing Suite that can be used to test system changes. By capturing the
workload in to two different STSs (for before and after change runs), one can use SPA
to assess the impact of the system change.
3.
With Oracle Enterprise Manager Cloud Control 12c, a “one-click” STS transport
mechanism can be used to simplify the process of moving STS workloads and tuning
artifacts such as SQL Profiles or Plan Baselines between production and test databases.
4.
SPA supports testing of databases consolidated through Oracle Multitenant Databases
or schema consolidation techniques.
6
Oracle Database 12c: Real Application Testing Overview
Database Replay
Database Replay provides DBAs and system administrators with the ability to faithfully,
accurately and realistically rerun actual production workloads, including online user and batch
workloads, in test environments. By capturing the full database workload from production
systems, including all concurrency, dependencies and timing, Database Replay enables you to
realistically test system changes by essentially recreating production workloads on the test system
– something that traditional testing tools that rely on simulation can never duplicate. With
Database Replay, DBAs and system administrators can test

Database upgrades, patches, parameter, schema changes, etc.

Configuration changes such as conversion from a single instance to RAC, ASM, etc.

Storage, network, interconnect changes

Operating system, hardware migrations, patches, upgrades, parameter changes

Database consolidation testing projects using Oracle Multitenant Databases and schema
consolidation

Workload scale-up and custom load testing scenarios
Lower test infrastructure cost
DBAs now have a test infrastructure at their disposal to test their changes without the overhead
of having to duplicate an entire application infrastructure. Database Replay does not require the
set up overhead of having to recreate a middle-tier or a web server tier. Thus, DBAs and system
administrators can rapidly test and upgrade data center infrastructure components with the
utmost confidence, knowing that the changes have truly been tested and validated using
production scenarios.
Faster deployment
Another major advantage of Database Replay is that it does not require the DBA to spend
months getting a functional knowledge of the application and developing test scripts. With a few
point and clicks, DBAs have a full production workload available at their fingertips to test and
rollout any change. This cuts down testing cycles from many months to days or weeks and brings
significant cost savings to businesses as a result.
Database Replay consists of four main steps as illustrated in Figure 2 and described below:
1.
Workload capture
When workload capture is enabled, all external client requests directed to the Oracle
Database are tracked and stored in binary files, called capture files, on the database server
host file system. Oracle recommends taking a backup of the entire database prior to the
workload capture. The user specifies the location of the capture files and the workload
7
Oracle Database 12c: Real Application Testing Overview
capture start and end time. During this process, all information pertaining to external
database calls is written to the capture files.
2.
Workload processing
Once the workload has been captured, the information in the capture files has to be
processed. This processing transforms the captured data into replay files and creates all
necessary metadata needed for replaying the workload. The capture files would typically be
copied to another system for processing. This must be done once for every captured
workload before they can be replayed. After the captured workload is processed, it can be
replayed repeatedly on a replay system. As workload processing can be time consuming and
resource intensive, it is generally recommended that this step be performed on the test
system where the workload will be replayed.
3. Workload replay
After the captured workload has been processed, it is now ready for replay. A client
program, called Replay Client, then processes the replay files and submits calls to the
database with the exact same timing and concurrency as in the capture system. Depending
on the captured workload, you may need one or more replay clients to properly replay the
workload. A calibration tool is provided to help determine the number of replay clients
needed for a workload. It should be noted that since the entire workload is replayed
including DML and SQL queries, it is important that the data in the replay system be
identical to that in the production system, whose workload was captured, to enable reliable
analysis for reporting purposes.
4. Analysis and Reporting
Extensive reports are provided to enable detailed analysis of the capture and replay. Any
errors encountered during replay are reported. Any divergence in rows returned by DML or
queries is shown. Basic performance comparisons between capture and replay are provided.
For advanced analysis, Replay Compare Period and other AWR reports are available to allow
detailed comparison of various statistics between capture and replays.
Both the workload capture and replay process support a filtering capability that is useful for
targeting workload of interest, such as by service, action, module to name a few. Oracle
Enterprise Manager Cloud Control 12c significantly enhances the value of Real Application
Testing by supporting end-to-end Database Replay automation. This simplifies the process of
saving and transferring the workload capture and performance data to test system, setting up the
test system and replay clients correctly, and orchestrating the entire replay through the Cloud
Control interface.
8
Oracle Database 12c: Real Application Testing Overview
Figure 2. Database Replay Workflow
Consolidated Database Replay
New in Oracle Database 12c, Database Replay supports simultaneous execution of multiple
database captures on a single consolidated database. The consolidated database can be a
Multitenant database or a traditional database consolidated using schema consolidation methods.
Replaying multiple workloads against a consolidated database gives assurance that the target
platform can support the workload. Database Replay supports captures from any supported
Oracle Database version. Database Replay can be executed on Oracle Database 11.1 and above.
Consolidated Database Replay can be executed on Oracle Database 11.2.0.2 and above. The
captures for Database Replay are platform agnostic and can be replayed on any supported
operating system.
9
Oracle Database 12c: Real Application Testing Overview
In addition, Consolidated Database Replay supports scheduling of the individual replays enabling
investigations of various workload scenarios.
Figure 3 shows the result of a Consolidated Replay of two workloads.
Figure 3. Consolidated Replay of Two Workloads
Database Replay Workload Scale-Up
Database Replay also supports creation of a new workloads based on existing captured
workloads. The new workloads can be used for capacity planning and validation of various
what-if workload scenarios. Three techniques that can be used with Database Replay to validate
consolidation include Workload Folding, Time Shifting and Schema Remapping.
10
Oracle Database 12c: Real Application Testing Overview
The first of these techniques is Workload Folding. Workload subsetting can be used to compose
new workloads. Existing captures can be divided into two smaller workloads by slicing an
existing captured workload into subsets by specifying a point in time within the captured
duration. Then you can double the workload by folding the workload along this specified pointin-time. This is done by submitting simultaneous replays of the subset workloads on the target
database. This effectively doubles the workload without the need to use scripting or supplying
binds. This technique is suitable for applications where individual transactions are mostly
independent of each other.
Another scale-up technique is Time Shifting. You can schedule multiple database replays so that
their peak database utilizations are aligned. This allows you to see if your target consolidated
system can handle the maximum production workload from your current production systems.
Oracle Enterprise Manger Cloud Control 12c Release 3, Plug-in Release 1 provides a
comprehensive and intuitive user interface for testing consolidation and capacity planning
through support for workload subsetting, workload scheduling, and multitenant architecture.
This enables businesses to future-proof their environment for growth and improve their SLAs.
Database Replay also supports testing with schema duplication. You can duplicate your target
schema and run multiple replays of the same workload. Before running these multiple replays,
you remap users so that each replay goes against its separate schema, avoiding workload
collisions. Schema duplication allows you to test multiple scales of the current workload,
maintaining the exact workload profile and concurrency. This is useful in scenarios such as
schema as a service (SaaS) or where each line of business has its own schema.
Workload Intelligence
Workload Intelligence enables you to use the information stored in database capture files in
additional ways, including the following:

Analyze and model the workload

Discover significant patterns and trends within the workload

Visualize what was running on the production system during workload capture
Workload Intelligence enables you to better visualize what a captured workload looks like by
exploring template patterns and the corresponding SQL statements. For each pattern, you can
view important statistics, such as the number of executions of a given pattern and the database
time consumed by the pattern during its execution. Some typical uses for Workload Intelligence
are workload modeling and profiling. Workload Intelligence provides context for application
optimization and performance tuning, identifies significant SQL statements, their patterns and
workload impact.
11
Oracle Database 12c: Real Application Testing Overview
Which testing solution
Choosing the right testing solution helps DBAs absorb and manage change efficiently.
SPA helps DBAs improve SQL response time. SPA is the SQL unit test to validate that all of
your individual SQL statements are performing optimally and should be used first in almost
every testing environment.
Database Replay is designed to test and improve overall system performance. After SPA has
validated the individual SQL performance, Database Replay should be used to ensure
performance under full system load.
Consolidated Database Replay provides the ability to assure desired database performance for
database consolidation projects, whether consolidating onto an Oracle Exadata machine, Oracle
Database Appliance, Oracle Multitenant Databases or other consolidated infrastructure.
Database Workload Scale-up and custom workload creation functionality enables you to futureproof your environment by making it possible to test the system under various what-if workload
scenarios.
Real Application Testing makes it easy for database administrators to manage and execute
changes that are critical to the business and do it all at lower risk.
Conclusion
Change is relentless in today’s rapidly evolving IT environments. But it doesn’t have to be
difficult for data center managers and administrators. Thanks to the Real Application Testing
capabilities in Oracle Database 12c, database administrators can adapt to changes easily while
eliminating any undesired side effects. . Numerous customers have already deployed Real
Application Testing and benefited from greatly reduced cost and effort in testing target
environments with enormous reductions in unplanned downtime. A Forrester research study
determined that customers who have deployed Real Application Testing have a 224% risk
adjusted ROI over a three year time frame with a payback period of 5.9 months.
Real Application Testing helps organizations lower their testing costs by giving DBAs and system
administrators an easy-to-deploy solution for testing and rolling out data center changes with
reduced hardware and software investments.
12
Oracle Database 12c:
Real Application Testing Overview
October 2013
Author: Jagan Athreya
Contributing Authors: Mughees Minhas, Kurt Engeleiter
Oracle Corporation
World Headquarters
500 Oracle Parkway
Redwood Shores, CA 94065
U.S.A.
Copyright © 2013, Oracle and/or its affiliates. All rights reserved. This document is provided for information purposes only and
the contents hereof are subject to change without notice. This document is not warranted to be error-free, nor subject to any
other warranties or conditions, whether expressed orally or implied in law, including implied warranties and conditions of
merchantability or fitness for a particular purpose. We specifically disclaim any liability with respect to this document and no
contractual obligations are formed either directly or indirectly by this document. This document may not be reproduced or
transmitted in any form or by any means, electronic or mechanical, for any purpose, without our prior written permission.
Worldwide Inquiries:
Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their
Phone: +1.650.506.7000
respective owners.
Fax: +1.650.506.7200
oracle.com
0109
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement