ABCs of the Autonomic Computing Toolkit

Introduction

Technology has permeated so many aspects of our life today that it is almost expected. Because technology appears in even the simplest tasks, it seems routine. However, with the evolution of a larger technical infrastructure comes the task of maintaining and managing it. Shouldn’t the very technology that makes life easier -- automates a set of business processes and integrates applications and data across a business -- be smart enough to manage itself? This guiding principle is the basis for the autonomic computing direction from IBM. Like the human nervous system, a certain level of management can be attained by making the system more aware of itself, and programming actions in response to certain situations.

Back to top

Autonomic computing overview

The IBM Autonomic Computing Toolkit is a collection of self-managing autonomic technologies that can provide this type of system awareness and control. The three main technology areas include problem determination, solution installation and deployment, and common systems administration. The problem determination tools and technologies are designed to reduce time spent in problem analysis and give a central point of interaction with multiple data sources. These technologies introduce standard interfaces and formats for logging and tracing, which are the key to building self-healing, self-optimizing autonomic computing system capabilities. The solution installation and deployment tools support automated tasks such as dependency checking and change management, making the process of installing software packages faster and less error-prone and, in the future, providing self-configuring capabilities. Last, but not least, the common systems administration function, based on the Integrated Solutions Console technology, supports a browser-based, common user interface that enables you to manage all systems in a consistent manner.

Problem Determination

Autonomic managers monitor managed resources and must be able to initiate actions in response to problems that arise that might affect normal operation. These actions can be defined as business policies, but the problem must first be identified from a potentially large set of messages. In addition, these messages might not even be in compatible formats. To address these issues, the problem determination architecture has standardized on a single message format: the Common Base Event. A Remote Agent Controller (RAC) that includes agents for multiple platforms to allow processing of remote log files using the Log Trace Analyzer tool is also included in the toolkit. Table 1 summarizes the problem determination components.

Solution Installation and Deployment

Solution installation and deployment is an important aspect of system administration. As more applications are deployed, the system becomes increasingly complex. One product might have several dependencies on other products. As the number of such dependencies increase, you face the challenge of balancing between versions of prerequisite software. The solution installation architecture is based on defining a common solution knowledge base to eliminate the ensuing prerequisite software complexity. This knowledge base is the basis for automated dependency checking, which is important in achieving a self-configuring system. Table 2 summarizes the solution installation components.

Common Systems Administration

The problem determination and solution installation technologies provide an infrastructure that helps to support the autonomic computing aspects of self-configuring, self-healing, self-optimizing, and self-protecting behaviors. However, no matter how automated the system becomes, at some point humans must interact with the system, especially in the early stages prior to attaining a completely autonomic system. You need to verify that as you progress, you are achieving the desired results. Of course, if you want humans to interact with the system, you need a user interface. Just as the problem determination and solution installation data was normalized or standardized, the systems administration console should also be consistent. The Autonomic Computing Toolkit promotes the use of the Integrated Systems Console technology as a common console technology for systems administration. It provides the main platform for hosting administrative functions as individual plug-ins to the console framework. The Integrated Solutions Console provides a consistent look-and-feel, thereby reducing the learning curve and adoption as new management components are introduced. Table 2 summarizes the common systems administration components.

Figure 1 shows an overview of the components of the Autonomic Computing Toolkit.

Back to top

What's new in Release 3?

Table 4 summarizes the additions and revisions in this release of the Autonomic Computing Toolkit.

Back to top

Installation

The Autonomic Computing Toolkit Web site contains many downloadable packages that can be installed and used to develop autonomic computing solutions. The Web site helps new users identify which of the packages are of interest and directs them to a specific download section to obtain the package.

The download section can also be accessed either directly from the navigation tab or from the Getting Started tab and following the links. The download section contains a link for each category and one for documentation. The Autonomic Computing Toolkit User’s Guide is particularly useful for new users. For advanced users, the Autonomic Computing Toolkit Developer’s Guide provides details on how to customize the sample scenarios or build your own autonomic capabilities. Finally, the example scenarios have individual guides (Autonomic Computing Toolkit Problem Determination Log/Trace Scenario Guide and the Autonomic Computing Toolkit Solution Installation and Deployment Scenario Guide), which explain how to install, configure, and run the scenarios to better visualize how the technologies can be used.

The packaging used by the Autonomic Computing Toolkit is based on the Solution Installation and Deployment technology, and checks that no other installations of a component or technology is present before installation. If a previous installation is present, the installation process stops and informs the user to remove the previous version before proceeding.

This article covers the installation of the entire Autonomic Computing Toolkit in a Windows™ environment; therefore, download each of the Windows components summarized in Tables 1 through 3. Note that some of the technologies have interdependencies or dependencies outside the Autonomic Computing Toolkit, as documented in Table 5.

Common systems administration

The Integrated Solutions Console, an IBM Self-Managing Technology Component, is used to build effective common systems administration capabilities. It is comprised of a run-time component, developer InfoCenter, sample components, and a toolkit plug-in component for developing custom component plug-ins for the Integrated Solutions Console. If you are going to use the toolkit plug-in to develop your own console components, you must install Rational Application Developer V6.0.

Note: The Integrated Solutions Console cannot coexist on a system that has WebSphere Portal installed. This includes previous versions of the Integrated Solutions Console, which contain WebSphere Portal. If the installation fails and an error message states that the Integrated Solutions Console is already installed, uninstall any previous versions of the Integrated Solutions Console or WebSphere Portal.

After installing the WebSphere Studio prerequisites:

1. Install the Integrated Solutions Console by executing the ISC_v6.0.0_win32.exe file.
2. Select OK to continue.
3. Select Next and accept the license agreement.
4. Specify a password for the Integrated Solutions Console administrator (iscadmin).
5. Accept the defaults for all the ports.
6. Leave the WebSphere Application Directory blank and install the embedded version with the Integrated Solutions Console.
7. Specify the home directory for WebSphere Studio to install the Integrated Solutions Console plug-in.
8. Accept the default for the Console Help Server port.
9. Specify the installation directory for the Integrated Solutions Console (or accept the default). Select Next.
10. Select Install to start the installation process. The process might take up to 60 minutes.
11. Select Finish when the installation process completes.

The Integrated Solutions Console requires the ISC_Portal server to be running. The server remains started after installation; however, if the system is rebooted, the ISC_Portal server will no longer be running. If you need to start the console, execute the shortcut: Programs > IBM Autonomic Computing Toolkit > ISC > Start ISC Server. To verify that the Integrated Solutions Console was successfully installed:

1. Open the Integrated Solutions Console by pointing to the following URL:

   http://your.server.name:isc_port/ibm/console

   where your.server.name is the host name for the Integrated Solutions Console installation and isc_port (8421 by default) is the port for Integrated Solutions Console that you specified during the installation process. You must specify the protocol name (http) in the URL because the URL contains a port number.
2. Log in as the Integrated Solutions Console administrator. Specify the user ID that was configured during the installation process. The default user ID is iscadmin.
3. Select the Settings tab. The Settings navigation tree is displayed.
4. Select User and Group Management. The page is displayed in the work area.

   To view help for the portlet on the page, click the portlet help icon (the ? symbol). The Manage users and groups help topic is displayed in a separate browser window (Figure 3).

5. On the Integrated Solutions Console toolbar, select Help. A separate window is launched and displays a navigation frame for accessing all of the Integrated Solutions Console Basics help and the Integrated Solutions Console Developer InfoCenter.
6. To log out of the console, select Log off on the toolbar. The Login page is displayed.
7. To stop the Integrated Solutions Console, execute the shortcut: Programs > IBM Autonomic Computing Toolkit > ISC > Stop ISC Server

Customizing the Integrated Solutions Console
To develop custom Integrated Solutions Console components that perform administrative functions, refer to the Create an administrative suite tutorial. To develop a help system for the Integrated Solutions Console, refer to the "Enable a help system within the Integrated Solutions Console" article. To debug a console plug-in remotely, refer to the "Enable WebSphere Studio Application Developer for remote debugging of Integrated Solutions Console components" article.

Problem Determination

It is recommended that the Integrated Solutions Console be installed first if you want to execute and observe the Problem Determination scenario. After the Integrated Solutions Console is installed, the remaining prerequisite technologies can be installed. Because the Problem Determination scenario highlights the Common Base Events being passed from the GLA to AME using Web services, the Embedded WebSphere Application Server - Express bundle must be installed next, as it is a prerequisite of both GLA and AME.

Now that the Integrated Solutions Console is installed, you can install the Embedded WebSphere Application Server - Express, the Autonomic Management Engine, and the Generic Log Adapter. These are all prerequisites for the Problem Determination Scenario, which can be installed next.

Embedded WebSphere Application Server - Express

To install the Embedded WebSphere Application Server - Express bundle:

1. Execute the WebSphereX_v6-0-0-2_win32.exe (for Windows) file. A pop-up window opens stating that WebSphere Version 6.0.0.2 is being installed on your system.
2. Select Next to proceed to the license agreement panel.
3. Select Next to accept the terms of the license agreement.
4. The appropriate system checks are performed, and a panel indicates all the ports that will be used by the application server. Accept the defaults, or enter the port values required by the system and select Next.
5. Choose the installation directory or retain the default directory and select Next.
6. A summary panel is shown indicating the settings that will be used for the installation. Select Install to start the installation process.
7. Upon successful completion of the installation process, a final panel appears indicating the completion status and an option to view the README file before exiting. Select Finish to complete the installation wizard.

This installs a Web services stack that can be used by both the Autonomic Management Engine and the Generic Log Adapter.

Now that the prerequisite for Web services is installed, you can install the Autonomic Management Engine and the Generic Log Adapter.

Autonomic Management Engine

The Autonomic Management Engine is a fully functional example implementation of an autonomic manager. It uses a resource model to provide customized behavior.

To install the Autonomic Management Engine in Web services mode:

1. Execute the AME_v1.2.0_win32.exe (for Windows) file. A pop-up window opens stating that AME Version 1.2.0 is being installed on your system.
2. Select Next to proceed to the license agreement panel.
3. Select Next to accept the terms of the license agreement.
4. The next panel requests the type of AME installation. Select Install and deploy AME as a Web service because that is required by the Problem Determination scenario, and select Next to proceed to the installation location panel.
5. Choose the installation directory or retain the default directory and select Next.
6. A summary panel is shown indicating the settings that will be used for the installation. Select Install to start the installation process.
7. Upon successful completion of the installation, a final panel appears indicating the completion status and an option to view the README file before exiting. Select Finish to complete the installation wizard.

To verify the AME installation, use the Simple Agent Reference Application (SARA), which is provided with the AME. SARA can run in either stand-alone mode or Web services mode. Refer to the Autonomic Management Engine 1.2 Developer’s Guide in the <AME_install_dir>\docs directory.

1. Customize the following variables in the sara.bat file (for Windows) located in the <AME_install_dir>\sara directory:
   SET ARCH=w32-ix86 (set based on the system you are running on)
   SET SARA_TOP=C:\AUTONO~1\ (should reference the AME installation directory)
   SET JAVA_HOME=C:\PROGRA~1\IBM\Java142\jre (should contain the Java home path to use)
   
   
2. Start the SARA application in Web services mode. This mode requires that the web server be started first. Then start SARA on the command line using the sarawscli.bat file.
3. At the ready prompt, type help for the application syntax.

Figure 4 shows the SARA command line and syntax.

4. Start the application and underlying AME (use the -reset option if it is already started).
   

   startrme

5. Now that the engine is started, you need to load one or more resource types. There are samples provided in the <AME_install_dir>\samples directory: JVM Monitor and Folder Monitor representing CIM resource models, and Folder_Monitor_CBE and Sample_Monitor_CBE representing CMM resource models.
   

   instrmtype ..\docs\samples\Folder_Monitor_CBE.zip

6. List the resource types.
   

   lsrmtypes

7. Create a resource model instance (fm) of the Folder Monitor. The resource model instance is the function that will monitor and interact with the resource.
   

   mkrminstance fm Folder_Monitor_CBE

8. Start the resource model instance (fm) to asynchronously run the decision algorithm for the CMM manageability model.
   

   startrminstance fm

9. List the resource model instances.
   

   lsrminstances

10. Stop each resource model instance.
    

    stoprminstance fm

11. Exit the program.
    

    quit

Using the Autonomic Management Engine
To embed AME into one of your own applications, refer to the Autonomic Management Engine 1.2 Developer’s Guide and the Problem Determination Scenario available in the Autonomic Computing Toolkit.

Resource Model Builder

The Resource Model Builder is a tool that lets a user build a model for the resource that is to be managed by the AME. To install the RMB:

1. Install the prerequisite Eclipse 3.0 bundle included in the Autonomic Computing Toolkit. The installation process creates a directory called <Eclipse_install_dir>\DevelopmentKit.
2. Install the RMB bundle by unzipping the RMB_v1-2-0_win32.zip bundle directly into the <Eclipse_install_dir>\DevelopmentKit directory and accept all file overwrites. The RMB bundle is a set of plug-ins that install on an Eclipse framework.

To verify the RMB installation, start the Eclipse application that includes the RMB:

1. Start the Eclipse environment by selecting Programs > IBM Autonomic Computing Toolkit > Developmentkit > Start Eclipse.
2. Specify the workspace location (or accept the default) and select OK. The Eclipse platform opens.
3. Close the Eclipse welcome page to enter the workbench.
4. Create a new project.
   1. Select File from the main menu or right-click in the Navigator pane on the left.
   2. Select New > Project > IBM Tivoli Project and expand that project folder.
   3. Select CMM Resource Model Project to create a new CMM resource model project. Click Next.
   4. Specify the project name and select Finish.
   5. Accept the prompt to open the IBM Tivoli Management Perspective by clicking Yes.
5. Create a Resource Model in the new project.
   1. Select the new project node in the Navigator pane and right-click.
   2. Select New > CMM Resource Model to create a basic CMM resource model. Alternatively, you can create an empty resource model, which has its own wizard.
   3. Specify a new CMM resource model project name and select Next.
   4. Define the general settings for the new resource model by specifying the Internal name, version, Algorithm class name, Description, and supported platforms the resource model will run on. Click Next.
   5. Specify the identifier for each managed resource monitored by this resource model. The resource model will process only those incoming Common Base Events that match the specified managed resource identifiers (MRID). Common Base Event filters can also be created and associated to each MRID by selecting the Add button in the filter section of the panel.
   6. Select Finish to complete the resource model.
   7. Expand the project folders to view the model components.

Creating a custom resource model

To create a simple resource model for processing Common Base Events asynchronously, refer to the Autonomic Management Engine 1.2 Developer’s Guide and the Problem Determination Scenario included in the Autonomic Computing Toolkit. For details on how to create a Tivoli management console resource model, refer to the IBM Tivoli Monitoring Version 5.1.1 Creating Resource Models and Providers redbook.

Generic Log Adapter Runtime and Rule Sets
The Generic Log Adapter run time is needed to produce Common Base Events from log files based on user-written rules. To install the Generic Log Adapter Runtime and Rule Sets:

1. Execute the GLA_v3-3-0_win32.exe file. A pop-up window opens stating that Generic Log Adapter version 3.3.0 is being installed on your system. Select OK to continue.
2. Select Next to accept the terms of the license agreement.
3. The next panel requests the type of GLA installation. Select Install and deploy GLA Subscriber to allow the GLA to send Common Base Events using Web services in the Problem Determination Scenario. Select Next to proceed to the installation location panel.
4. Choose the installation directory or retain the default directory and select Next.
5. A summary panel is shown indicating the settings that will be used for installation. Select Install to start the installation process.
6. Select Finish when the installation completes.

To verify the installation, run one of the sample adapter configurations:

1. Edit the runregex_example.bat file in the <GLA_install_dir>\config\IHS\access\v1.3.19.3 directory. Specify the GLA_HOME value, which is where you installed GLA (the <GLA_install_dir>).

   SET GLA_HOME=C:\Program Files\IBM\AutonomicComputingToolkit\GenericLogAdapter

2. Verify that the JAVA_HOME environment variable is set and the PATH variable includes the %JAVA_HOME%\bin value.
3. Run the runregex_example.bat file.
4. You should see an example.out file with the generated Common Base Events.

For additional information, refer to the Generic Log Adapter Getting Started Guide and the readme in the <GLA_install_dir>\docs directory.

Generic Log Adapter/Log and Trace Analyzer Plug-ins

To install the Generic Log Adapter and Log and Trace Analyzer tools:

1. You can choose to reuse the existing Eclipse environment installed for the RMB tool or install a separate one for the GLA/LTA plug-ins. The existing Eclipse installation is used here.
2. Install the GLA-LTA bundle by unzipping the GLA-LTA_v3-3-0_win32.zip bundle directly into the < Eclipse_install_dir >\DevelopmentKit directory and accept all file overwrites. The GLA-LTA bundle is a set of plug-ins that install on an Eclipse framework.

To verify the Log and Trace Analyzer installation:

1. Start the Eclipse environment by selecting Programs > IBM Autonomic Computing Toolkit > Developmentkit > Start Eclipse.
2. Specify the workspace location and select OK.
3. Create a new empty project.
   1. Select File from the main menu or right-click in the Navigator pane on the left.
   2. Select New > Project > Simple > Project. Click Next.
   3. Specify the project name and select Finish.
   4. Add the example file used by the GLA verification above by selecting the project node and right-click.
   5. Select Import > File System and then select Next.
   6. Browse to the <GLA_install_dir>\config\IHS\access\v1.3.19.3 directory and select the regex_example.adapter file and then select Finish.
4. Double-click on the file and open it in the adapter editor.
5. Expand the adapter configuration contents and browse.

Creating custom log adapters

To create a custom log adapter, refer to the "Using Java class callouts with the Generic Log Adapter" article and the Using the Generic Log Adapter with the Log and Trace Analyzer, Create GLA components, and Create GLA sensors and outputters tutorials. To create a custom correlation engine, refer to the Create a Correlation Engine for the Log and Trace Analyzer article.

Agent Controller

The Agent Controller bundle contains the Remote Agent Controller for multiple platforms to allow processing of remote log files using the Log and Trace Analyzer tool. To install the Agent Controller package on a remote workstation:

1. Install the RAC bundle by unzipping the RACsecure_v6-0-1.zip content into an installation directory.
2. Extracting the RAC bundle creates installer directories for each of the supported platforms. For Windows, go to <RAC_install_dir>\IBM_AgentController\installers\win_ia32 and run the setup.exe program.
3. Select Next on the welcome screen to proceed.
4. Close all Eclipse platforms before continuing with the installation. Select Next to proceed to the license agreement panel.
5. Read and accept the license agreement and then select Next to proceed to the installation location panel.
6. Specify the location where the Agent Controller should be installed and select Next.
7. Select the Agent Controller features you would like to install:
   • Hyades Data Collection Engine − a data collection framework provided by the Eclipse open source project
   • Logging − an event-based logging framework for profiling and tracing applications
   • Component Test − legacy support for manual test enablement
   • Runtime Analysis
   • Security
   • Remote support for WebSphere Application Server − function for connecting to a remote
8. Select Next.
9. Specify the path of the Java Runtime executable program java.exe that the Agent Controller should be using. To use the Java Runtime Environment (JRE) provided by the Autonomic Computing Toolkit, choose C:\Program Files\IBM\Java142\jre\bin\java.exe. Select Next.
   
10. Optionally, specify the paths of IBM WebSphere Application Server Version 5.0 and 5.1, if installed. Select Next. Note that Version 6.0 has a built-in feature to handle this deployment function, so there is no need to specify this path.
11. Specify the hosts that can access the Agent Controller. Select Next.
12. Specify the security setting as Enabled (encrypted with authentication) or Disabled (no encryption). Select Next to proceed to the summary window.
13. Select Next on the summary screen to initiate the installation.
14. Select Finish to complete the installation.

To verify the Agent Controller installation:

1. On Windows, the agent is installed as a service and should start automatically.
2. Open the Windows Services dialog, and the IBM Rational Agent Controller should be started.
3. Stop and restart it to test the service.

Problem Determination Scenario

Now that all the prerequisites are installed, install the Problem Determination (PD) Scenario:

1. Execute the PDScenario_v3-0-0_win32.exe file. A panel opens stating that Problem Determination Scenario version 3.0.0 is being installed on your system. Select OK to continue.
2. Select OK.
3. Read and accept the license agreement and then select Next.
4. Select the installation type as Install files and setup the scenario to install the bundle so that it is operational with the other prerequisite already installed. Select Next to continue.
5. The next panel lists the required prerequisites for the PD scenario. All prerequisites have been installed, click Next.

   Note that the PD Scenario can be installed without the prerequisites, but it will not be functional. This option allows the user to view the implementation code and documentation.

6. Specify the location where the PD scenario should be installed (or retain the default) and select Next.
7. Specify the port to be used by the Cloudscape database (or retain the default) and then select Next.
8. Optionally set up the scenario to send events to the Tivoli Enterprise Console. An installation of Tivoli Enterprise Console is required to demonstrate this option. Select Next to open the installation summary window.
9. Select Install to start the installation process.
10. The next panel prompts for the Integrated Solutions Console administrator User ID and password. The PD Scenario deploys a component plug-in to the Integrated Solutions Console to control and monitor the scenario. Enter the user ID and password and select Next.
11. Select Finish when the installation completes.

To verify the installation, run the Problem Determination Log/Trace Scenario in the Example Scenarios section.

For more information on the concept of autonomic problem determination and the role it plays in achieving a self-healing system, refer to the "Automating problem determination: A first step toward self-healing computing systems" article.

Solution Installation and Deployment

The Solution Installation and Deployment support includes three scenarios. Two of the scenarios illustrate products from IBM business partner Macrovision. FLEXnet Publisher Installation Module (FNPIM) and InstallAnywhere (IA) implementations are demonstrated. The Samples scenario includes two examples, one that shows how to execute a Java custom action and another that illustrates a full software life cycle scenario.

The Integrated Solutions Console is an optional prerequisite for operating the Samples scenario. Command line operation of the Samples scenario does not require that the Integrated Solutions Console be installed. To install the samples scenario to use the Integrated Solutions Console:

1. Execute the SI_Samples_v3-0-0_win32.exe file. A panel opens stating that Solution Installation Samples Scenario version 3.0.0 is being installed on your system. Click Next to continue.
2. Read and accept the license agreement and then select Next to proceed to the installation type window.
3. Select Install files and setup the scenario on ISC install type and select Next.
4. Specify the installation folder destination and select Next.
5. Verify the content of the summary panel and select Install to start the installation process.
6. The next panel prompts for the Integrated Solutions Console administrator User ID and password. The Samples Scenario deploys a component plug-in to the Integrated Solutions Console to control and monitor the scenario. Enter the user ID and password and select Next.
7. Select Finish to complete the installation.

To verify the installation, run the Solution Installation Samples scenario.

FNPIM Solution Installation Scenario

1. Execute the SI_FNScenario_v2-0-0_win32.exe (for Windows) file.
2. When the installation screen appears, select Next and accept the license agreement.
3. Specify the installation directory for the scenario files and select Next.
4. Select Finish when the installation completes.

To verify the installation, run the Solution Installation with FNPIM in the Example Scenarios section.

InstallAnywhere Solution Installation Scenario

1. Execute the SI_IAScenario_v3-0-0_win32.exe (for Windows) file.
2. When the installation screen appears, select Next and accept the license agreement.
3. Specify the installation directory for the scenario files and select Next.
4. Select Finish when the installation completes.

To verify the installation, run the Solution Installation with InstallAnywhere in the Example Scenarios section.

Back to top

Example Scenarios

Problem Determination Log/Trace Scenario

The Problem Determination scenario represents a simple self-managing system that uses an intelligent control loop to collect system information, analyze it, plan appropriate responses, and then make necessary adjustments to resolve problems. This scenario exposes the specific technologies that make up a realistic self-healing system, and demonstrates how they work together to achieve an adaptive level (level 4 of the autonomic maturity levels) of self management. Figure 8 summarizes the problem determination scenario architecture.

1. Start the Integrated Solutions Console servers by selecting Start > Programs > IBM Autonomic Computing Toolkit > Problem Determination Scenario > Start PDScenario Servers. This starts the WebSphere Application Server and Integrated Solutions Console portal servers. The Cloudscape™ database, the Web application that uses Cloudscape, the GLA, and AME are started later when the scenario is invoked. Refer to the Problem Determination Log/Trace Scenario Guide for details.
2. Open the Web console to control the scenario by selecting Start > Programs > IBM Autonomic Computing Toolkit > Problem Determination Scenario > Launch Web Console and log into the Integrated Solutions Console as the administrator user (iscadmin).
3. Expand the menu item Autonomic Computing Scenarios in the Navigation Area window and view the PD scenario description (Autonomic Computing Scenarios > Problem Determination Scenario > Problem Determination Scenario Description).

4. Access the scenario control panel by selecting Autonomic Computing Scenarios > Problem Determination Scenario > Scenario Control.
5. Start the scenario by selecting the Begin Scenario link in the Problem Determination Scenario Control Panel.
6. View the status as the scenario components are initialized in the Problem Determination Scenario Status panel.
7. After the scenario has been set up (status is complete), view the PD Web Application by selecting the link in the Problem Determination Scenario Control panel.
   
8. The PD Web Application displays a set of table entries that contain information retrieved from the Cloudscape sample database. This data is refreshed at a regular interval to represent the operational state of the managed resources.
9. The user has control of when to induce an error in the operational control loop. To induce an error condition, select the Induce Condition link in the Problem Determination Scenario Control panel. This creates a problem in the application that causes the PD Web Application to stop displaying valid data. The reason for the error is not apparent.

   Because the PD Web Application can no longer display valid data, an HTTP error page is displayed.

10. The induced error is stopping the Cloudscape database. View the status as the scenario components are affected in the Problem Determination Scenario Status window. The autonomic manager is monitoring for this type of error and upon detecting the Cloudscape database being down, issues a restart of the Cloudscape database.
11. After the database restarts, the PD Web Application browser window refreshes and shows the data tables again, indicating a steady state has been restored.
12. The Log View window contains links to the product activity log file and the translated Common Base Event formatted log file.
13. End the scenario by selecting the Close Scenario link in the Problem Determination Scenario Control Panel.
14. Stop the scenario servers by selecting Start > Programs > IBM Autonomic Computing Toolkit > Problem Determination Scenario > Stop PDScenario Servers.

Solution Installation Samples

There are two solution installation samples included in this component: the Java Custom Action example and the Family Full Lifecycle example. The first sample demonstrates the use of the Java Custom Action feature of the solution installation and deployment technology. This sample shows how to create a new directory with a text file as part of a typical installation. But this is only one possible custom action that can be performed using this feature. The second sample illustrates the full software life cycle of a solution offering for developers and software package authors utilizing the solution installation and deployment technologies. It demonstrates the following typical life cycle actions:

• Install a base package
• Perform an initial configuration
• Apply a fix to the base package
• Undo a fix to the base package
• Apply an incremental update to the base package
• Perform a migration of the package after the incremental update
• Apply a full update to the package
• Uninstall the entire package offering

This scenario demonstrates a typical full life cycle of a software offering based on family members. The software versions, fixes, and updates are represented as text files within a directory structure named for members of a family. Figure 11 summarizes the four packages in the entire scenario.

To execute the solution installation scenarios:

1. Start the Integrated Solutions Console servers by selecting Start > Programs > IBM Autonomic Computing Toolkit > Solution Install Samples > Start ISC Servers. This starts the WebSphere Application and Integrated Solutions Console portal servers. Refer to the Solution Installation and Deployment Scenario Guide for details.
2. Open the Web console by selecting Start > Programs > IBM Autonomic Computing Toolkit > Solution Install Samples > Launch Web Console, and log in to the Integrated Solutions Console as the administrative user (iscadmin).
3. Expand the menu item Autonomic Computing Scenarios in the Navigation Area window and view the scenario description by selecting Autonomic Computing Scenarios > Solution Install and Deployment Samples > Solution Install Scenario Description.

4. To view the individual sample descriptions, select the sample description links for the Java Custom Action and Family Lifecycle samples (that is, expand the Autonomic Computing Scenarios > Solution Install and Deployment Scenarios > Samples > Java Custom Action > Sample Description and Autonomic Computing Scenarios > Solution Install and Deployment Scenarios > Samples > Family Lifecycle > Sample Description, respectively).
5. To execute the Java Custom Action sample:
   1. Navigate to the Control Panel by selecting Autonomic Computing Scenarios > Solution Install and Deployment Scenarios > Samples > Java Custom Action > Control Panel.
   2. Browse the solution install sample descriptors by selecting the packagedIU.xml and MyCustomActionArtifact.xml links in the descriptor portlet.
   3. Select the Install Sample link in the Solution Install Control Panel.
   
   Figure 13. Java Custom Action Solution Install Scenario control panel
   
   
   4. When the installation completes, the Un-install Sample link becomes active. Also, a directory is created by the Java custom action (<SI_SCENARIO_INSTALL_DIR>/samples/javacustomaction/jca_test) and a file, called jca_test.txt, is created in that new directory. Verify that this directory and file exist.
   5. To uninstall the sample, select the Un-install Sample link in the control panel. After the uninstall completes, manually delete the jca_test directory and file because an uninstall artifact to delete them was not included in the sample.
6. To execute the Family Lifecycle sample:
   1. Navigate to the Control Panel by selecting Autonomic Computing Scenarios > Solution Install and Deployment Scenarios > Samples > Family Full Lifecycle > Control Panel.
   2. Browse the solution install sample descriptors by selecting the base/packagedIU.xml, fix/packagedIU.xml, incremental/packagedIU.xml, and fullupdate/packagedIU.xml links in the descriptor portlet.
   3. Select the Install V1.2.0 link in the Solution Install Sample Control Panel. This installs the base package, which consists of the following directory structure off of the <SI_SCENARIO_INSTALL_DIR >/samples/family/installed directory. The Configure base link becomes enabled. Each base directory contains a text file with the string Version 1.2.0.
   
   Figure 14. Lifecycle Solution Install Scenario control panel
   
   
   4. Select the Configure base link in the Solution Install Control Panel. This starts the configuration process. When the configuration has completed, the Apply fix and Apply incremental update links become active. The user has the option to apply either or both of these installations. Each text file should now include Config Complete.
   5. Select the Apply fix link to install the fix. When the fix installation has completed, the Rollback fix link becomes active. A second text file should now be visible in the daughter directory (daughter_1.2.0_Fix.txt), with the text Version 1.2.0.
   6. Select the Rollback fix link to uninstall the fix representing the removal of a bad fix.
   7. Apply the incremental update by selecting the Apply incremental update link. When the incremental update has completed, a new daughter text file (daughter_1.5.0) is in the daughter directory, and the Migrate update link becomes active.
   8. Select the Migrate update link to migrate the incremental update to the base installation. When the migration completes, the Un-install V1.2.0 link is enabled, as well as the Install full update V2.0.0 link. If the Un-install V1.2.0 link is selected, the sample is removed and the scenario is completed. If the Install full update V2.0.0 link is selected, the system is upgraded to Version 2.0.0. The daughter 1.5.0 text file should now contain the text Migrate Complete because the fix only applies to children who have migrated from Version 1.2.0.
   9. Initiate the full update by selecting the Install full update V2.0.0 link. The update is installed on top of the Version 1.2.0 base. When the update completes, the Migrate to V2.0.0 and Un-install V2.0.0 links become active. A new set of V2.0.0 text files are created (Dad_2.0.0.txt, Mom_2.0.0.txt, and daughter_2.0.0.txt) and now contains the text Version 2.0.0.
   10. Select the Migrate link to move to the V2.0.0 update. All the V2.0.0 text files now contain the text Migrate Complete, and the Un-install V2.0.0 link should be active.
   11. Select the Un-install V2.0.0 link to start removing all the family packages. All the children and parent nodes are removed and the family/installed directory should now be empty. The Install V1.2.0 link should be the only active link. This is used to start the process of installing the base version again.
7. Log out of the Integrated Solutions Console and close the browser.
8. Stop the Integrated Solutions Console servers by selecting Start > Programs > IBM Autonomic Computing Toolkit > Solution Install Samples > Stop ISC Servers.

Solution Installation with FNPIM Scenario

To install the FNPIM Solution Installation example:

1. Execute the SI_FNScenario_v2-0-0_win32.exe (for Windows) file.
2. Select OK to continue.
3. Select Next and accept the license agreement.
4. Specify the installation type preferred (use the default value Typical).
5. Specify the installation directory for the scenario files and select Next.
6. After the installation completes, select Finish.

To verify the installation:

1. Up to four IUs (Installable Units) are installed under this scenario, depending on the features selected during the installation process. The four IUs are:
   
   • FNPIM Demo - An HTML file that loads a Macromedia demo describing FLEXnet Publisher Installation Module (FNPIM), the product used to package this scenario. It is Macrovision's innovative new development environment for creating software packages in the solution installation and deployment format.
   • Galatea - A single file, Galatea.txt.
   • Proteus - A single file, Proteus.txt.
   • Triton - A single file, Triton.txt.
2. When installing Galatea, Proteus, or Triton, the files are installed into the target directory.
3. When installing the FNPIM demo, two files are copied: index.html and installx2.swf. To look at the Flash demo, start a Web browser and open the index.html file. The Flash demo starts automatically if you have a Flash browser plug-in installed.

Solution Installation with InstallAnywhere Scenario

To install the InstallAnywhere Solution Installation example:

1. Execute the SI_IAScenario_v3-0-0_win32.exe (for Windows) file.
2. Select the JVM to use for the installed application (use the JRE 1.4.2 installed previously).
3. Select Install, and then select Next.
4. Specify the installation directory for the scenario files and select Next.
5. Select Next if the requirements pass.
6. Select Install to start the installation process.
7. After the installation completes, select Done to exit the installer.

To verify the installation:

1. Four IUs are installed under this scenario. The four IUs are:
   
   • Documentation - A readme.html file and the .zip file containing the actual package that was used during the installation. You can extract the packagedIU.xml file in the .zip archive to look at the solution installation descriptor.
   • Galatea - A single file, Galatea.txt.
   • Proteus - A single file, Proteus.txt.
   • Triton - A single file, Triton.txt.
2. When installing Galatea, Proteus, or Triton, the files are installed into the target directory.
3. When installing Documentation, three files are copied: readme.html, style.css, and NeptuneMoonsPackage.zip. To look at the readme file, open a Web browser and open the readme.html file. To look at the solution installation descriptor, use an unzip program to extract the packagedIU.xml file. Use any XML editor or browser to view the descriptor.

Back to top

Glossary

Autonomic Computing (AC) – striving to achieve a self-configuring, self-healing, self-optimizing and self-protecting (that is, self-managing) environment in order to reduce the overall total cost of ownership for new technology.

Autonomic Computing Toolkit (AC Toolkit) - a set of technologies that can be leveraged to provide autonomic system awareness and control.

Autonomic Management Engine (AME) – an example of an autonomic manager.

Autonomic Manager – this is the component that does the monitoring and controlling of managed resources.

Common Base Event – the common log event format that is used to normalize or standardize the source of problem information across a managed system of resources.

Common Systems Administration (CSA) - based on the Integrated Solutions Console technology, supports a browser-based common user interface that enables you to manage all systems in a consistent manner.

Generic Log Adapter (GLA) – simplifies adoption of the Common Base Event. It provides an editor that maps log content to Common Base Event attributes. It supports an embedded test environment for the transform and integrates with the Log and Trace Analyzer tools.

Integrated Solutions Console – a console framework based on WebSphere Portal that provides a consistent look and feel for system administration.

Log and Trace Analyzer (LTA) – provides an example of tooling for log browsing, analysis, correlation, and symptom database in support of problem determination.

Managed Resource – the component that is being monitored and controlled by the management system. It can be a single resource or a collection of resources.

Problem Determination (PD) – the act of analyzing problems in order to inform the manager to act accordingly to ensure normal system operation. This might include correlation of multiple information resources such as determining a root cause for the problem.

Solution Installation and Deployment – supports the process of installing software packages, making it faster and less error-prone and, in the future, provide self-configuring capabilities.

Back to top

Technical Support

For technical support, refer to the Autonomic Computing Forum.

Back to top

Acknowledgement

The author wishes to thank Lori Small, the original author of this article, for the article idea and for her initiative in getting it off the ground. Lori Small is a member of the SMB Emerging Technologies group, where she is researching how development and operational tools are used in the mid-market of SMB. You can reach Lori at lsmall@us.ibm.com.

Resources

• A Practical Guide to the IBM Autonomic Computing Toolkit (April 2004), by Jacob et. al. This redbook provides a view of the Autonomic Computing Toolkit as it exists today and describes how it can be used to enable autonomic capabilities.
  
  
• IBM Tivoli Monitoring Version 5.1.1 Creating Resource Models and Providers (August 2003), by Bhe et. al. This redbook focuses on using the IBM Tivoli Monitoring Workbench to build resource models that use existing ILT providers and custom ILT Java providers.
  
  
• "An autonomic computing roadmap" (developerWorks, February 2004) provides a roadmap to begin integrating autonomic computing concepts into your products.
  
  
• "Understand the autonomic manager concept" (developerWorks, February 2004) presents both a basic and a more detailed scenario of a working autonomic manager along with examples of IBM software applications that have autonomic functions.
  
  
• "Understand autonomic maturity levels" (developerWorks, February 2004) introduces and details the autonomic maturity levels and the evolution of the maturation process.
  
  
• "Take a quick tour of autonomic computing" (developerWorks, April 2004) explains the concepts behind autonomic computing and looks at the tools at your disposal for making it happen.
  
  
• "Simplify deployment tasks with Solution Installation technology" (developerWorks, May 2005) discusses the details of working with solution installation and deployment packages.
  
  
• "Understand problem determination" (developerWorks, March 2004) looks at the fundamental issue of problem determination and what it takes to transform this from a human process to an autonomic one.
  
  
• "Create an administrative suite" (developerWorks, February 2004) reviews the process for creating an Integrated Solutions Console component that performs administrative functions.
  
  
• "Enable a help system within the Integrated Solutions Console" (developerWorks, May 2004) discusses the complete cycle of creating a help plug-in for the Integrated Solutions Console.
  
  
• "Enable WebSphere Studio Application Developer for remote debugging of Integrated Solutions Console components" (developerWorks, September 2004) discusses how to debug Integrated Solutions Console components using WebSphere Studio Application Developer and WebSphere Portal Toolkit.
  
  
• "Automating problem determination: A first step toward self-healing computing systems" discusses the concept of autonomic problem determination and the role it plays in a achieving a self-healing system.
  
  
• "Understand the Autonomic Management Engine" (developerWorks, June 2004) discusses how AME works and shows you how to embed it into one of your own applications.
  
  
• "Create a simple resource model for processing Common Base Events from a file" (developerWorks, June 2004) describes how to create a simple resource model for processing Common Base Events that are stored in a file.
  
  
• "Using the Generic Log Adapter with the Log and Trace Analyzer" (developerWorks, June 2004) introduces the generic log adapter and log trace analyzer components.
  
  
• "Create GLA components" (developerWorks, July 2004) discusses how to create generic log adapter components.
  
  
• "Create GLA sensors and outputters" (developerWorks, September 2004) illustrates how to develop generic log adapter sensors and outputters.
  
  
• "Create a Correlation Engine for the Log and Trace Analyzer" (developerWorks, September 2004) illustrates the procedure for developing a custom correlation engine as a plug-in for the Log and Trace Analyzer.
  
  
• "Create a resource model for use with the Autonomic Management Engine" (developerWorks, June 2004) demonstrates the basics of creating a model, testing a resource model, packaging the model, and deploying the model.
  
  
• The Autonomic Computing Toolkit User’s Guide provides installation and user information for the Autonomic Computing Toolkit.
  
  
• The Autonomic Computing Toolkit Developer’s Guide is a technical guide to assist programmers in using the technologies included in the Autonomic Computing Toolkit to develop autonomic solutions.
  
  
• The Autonomic Computing Toolkit Problem Determination Log/Trace Scenario Guide gives you an introduction to and installation instructions for the Problem Determination scenario.
  
  
• The Autonomic Computing Toolkit Solution Installation and Deployment Scenario Guide gives you an understanding of all three Solution Installation and Deployment scenarios.
  
  

About the author

Comments

Back to top

Trademarks

Table of contents

• Introduction
• Autonomic computing overview
• What's new in Release 3?
• Installation
• Example Scenarios
• Glossary
• Technical Support
• Acknowledgement
• Resources
• About the author
• Comments