From Eclipsepedia

COSMOS Main Page >

Contents 1 Overview 2 Setup 3 The Script 3.1 Working with SML 3.2 Browse IT Resources 4 The Messages

Overview

The objective of the COSMOS demo is to present an overview of the project, our progress to date, and explain the key value adds to the Systems Management community.

The presentation aims to demonstrate COSMOS within a 30 minute time limit. Although in general we try to avoid slides that are not tied to the demo, we do provide one chart for the project overview, and one for the roadmap & wrap up slide. That leaves about 20 minutes of demo-able material.

Since SML & CML is a new exciting development and is a fundamental part of the project, we concentrate of showing how SML & CML can be used in monitoring systems as an interface between data producers and data consumers in order to facilitate interoperability among operations management tools.

Setup

In the future, we will provide a zip file that includes everything you need (e.g. Tomcat, Dojo, Eclipse, COSMOS plug-ins, etc.) to conduct this demo, along with .bat files to run the various pieces. In the meantime, please follow the steps outlined on the End to End Sample page to setup the demo environment.

The Script

Working with SML

1. Start Eclipse by running startEclipse.bat in the cosmos-demo directory of the demo zip file, passing in the location of your Java 1.5 executable as an argument if necessary
2. Create a new project called "Data Center" by selecting File > New > Project... > General > Project.
3. Read the first 2 bullets on the "Working with SML" slide in the presentation
4. Go to File > Import > General > File System to import the SML-IF file datacenterSMLIFToImport.xml as a file to be managed in your Eclipse workspace
5. To import that same SML-IF file as resources into the file-based repository, right-click the file and choose Import... > Service Modeling Language > Interchange Format
6. Read the third bullet on the "Working with SML" slide
7. Edit the file Data Center\software\webserver\tomcat.xml. This file contains statistical data. Locate the <ctrl:keyset> element. Show that this represents the keyset value that would have to change if you were retrieving different data. Similarly, the datasource id is located in the name attribute of the <tomcat> root element of this file - changing this would cause the data visualization component to reference a different Tomcat server. The keyset and datasource values should match those entered in the assembly files used by the data collection component to collect data.
8. Edit the file Data Center\os\windows\WindowsServer1.xml. This file contains the CBE data. Again, locate the <ctrl:keyset> element; this identifies the resource to the data store. As before, to set the datasource id, change the name attribute of the root element, which in this case is <s:os.windows>.
9. Read the 4th bullet on the slide
10. Select the "Data Center" project, right-click and select Export... > Service Modeling Language > Interchange Format

Browse IT Resources

1. Go to next slide ("Browse IT Resources")
2. Start Tomcat by running startTomcat.bat in the cosmos-demo directory, optionally passing in the value for JAVA_HOME as the first argument
3. Read the first bullet on the slide
4. Run openSampleReport.bat to open the COSMOS reporting interface in your browser
5. Read the rest of the bullets on the slide and follow the instructions therein. Be sure you navigate to the two resources referenced in the previous section that contained keyset data, so the associated reports can be displayed.

Old demo is hidden following this - edit to view source

The Messages

By following the script we can deliver the following messages

(not necessarily in this order)

• COSMOS is a project that is starting to deliver some basic value
• We can tie together the full life cycle of systems management from development (BtM/ME) to production (COSMOS reports)
  • In the scenario, we can talk about how the current BtM tooling for CBE was used to generate the CBEs that get stored in the data store. We can say how the WSDM tooling was used to create the WEF events that get stored in there as well.
  • We can indicate how ME was used to configure/update the SML-IF documents that described the management capabilities of the system under test, e.g. if log or statistical data was avaialable.
  • We talk to the standard based control layer (stretch goal is to show it)
    • Note: We can turn on collection manually possibly if we wated to show live data. I think this should be a stretch goal once we get everything working. The progression would be:
      • Show the ez-bake oven approach to the data store.
      • Turn on data collection manually and show real event data
      • Use web pages to control the collection (SML-IF would play a part here)
  • We show re-use/collaboration between the projects (TPTP/COSMOS)
  • We show web-based reports (initially derived from the TPTP reports)
• We can indicate the importance of using a model based approach in systems management (and our model language is SML/CML)