Difference between revisions of "Higgins/Solutions"

Revision as of 10:57, 10 February 2007

Overview

A Deployment Scenario is a specific combination of Components that, when assembled and deployed result in an application or service that is identifiable to an end-user as a "whole" app or service. This page is intended to explain how to assemble building block Components into running apps and services. The indended audience is technical, but more about assembling, building and deploying, as opposed to "developing."

Some of the Deployment Scenarios are web services or webapps that have been deployed on Eclipse servers and can be used for testing and and development-related purposes. Examples would include a CardSpace-compatible IdP service (what Microsoft would call a "Managed Card Provider" (not to be confused with our use of the term provider)), or a MediaWiki app that supports OpenID sign-in, etc.

Each Deployment Scenario should be desribed in its own section on this page, and should include a table with rows that describe the Components involved. This table should include build scripts and other 3rd party libraries.

Deployment Scenarios

CardSpace-interoperable Identity Provider/STS

.

Paul's Sandbox (alternate table design for discussion purposes)

1) In this proposal a deployment scenario consists of one row per machine. Each row describes ONE computer's configuration (OS, runtime, link URL (e.g. WAG)). If a second computer is required, then a second row (e.g. for WOOF) is added.

2) We agree to a convention that CardSpace-interoperable IdP/STS contains these two sections:

• Overview (as you already have, except only for ONE machine (other machines would be on other rows))
• Assembly and Build instructions (new)

The Assembly and Build instructions section would start off with a bulleted list of components and other external stuff that you'll need. We can include links to the various required rows on the Components page tables as we've started doing.

Notice that "Assembly & Build" has been removed as a Column from the table.

The rows I've deleted (and moved into the above wiki page) were almost entirely redundant. .

Identity Agent (in-browser selector, hosted IdA)

• Requires the user to install the Higgins Browser Extension (HBX). HBX relies on a hosted Identity Agent (IdA) service

Identity Agent (rich client selector, hosted IdA)

• Requires the user to install the Higgins Browser Extension (HBX) and the ISS Client UI rich client card selector. ISS Client UI relies on a hosted Identity Agent service

Identity Agent (rich client selector, local IdA)

• Requires the user to install the Higgins Browser Extension (HBX) and the ISS Client UI rich client card selector. ISS Client UI relies on a local Identity Agent service.

I-Card Manager Enterprise Application

.

RPPS Web Application

.

RSS-SSE RP Test Application

.

Nightly Builds

Though certainly not a "deployment" in the usual sense, the Higgins project automatically builds some of the Components every night.

.

Conventions Used V2

Each Deployment Scenario is documented in a table

• The deployment owner is responsible for maintenance of this table
• Each computer involved has its own row. This row describes that computer's configuration (OS, runtime, etc.).

Table Columns

1. Deployment Scenario - link to wiki page describing deployment scenario (see "Deployment Description" section below)
2. OS - OS that this machine either (a) runs on (see URL column) or (b) has been tested on. Put in parens the OS number if more than OS instance is involved
3. Runtime - Runtime environment for this machine (e.g. JVM & version, Tomcat & version, etc.)
4. Binding - how will the service running on this machine be consumed
5. Open - open enhancements and bugs (Bugzilla) for this deployment scenario (Note: none are currently defined)
6. URL - endpoint that hosts a test version of the service (hosted by Eclipse Foundation)
7. Owner - person with overall responsibility for this deployment scenario (not individual components)

Example Table

Deployment Description

This wiki page must contain the following sections:

1. Overview --describes briefly the service running on this machine
2. Assembly and Build
   • A table with a row for each Higgins component required and a row for each non-Higgins component JAR, etc required. Each row has these columns:
     1. Name of component/jar
     2. Link to description --in the case of Higgins components and projects this should be a link to the Componentspage#<component-name> AND the row prefix (e.g. "A" or "B")
3. Deployment
   • Deployment "hints": things you might have had to do to "configure" the files. For example the STS and LDAP CP have several configuration files that need to be explained a little for anyone wanting to replicate the deployment to their own servers. Another example: Tomcat configuration
   • Any unusual steps related to configuring non-Higgins components

The Assembly and Build instructions section would start off with a bulleted list of components and other external stuff that you'll need. We can include links to the various required rows on the Components page tables as we've started doing.

Conventions Used V1

The tables on this wiki page have the following column structure:

1. Deployment Scenario - link to wiki page describing deployment scenario
2. OS - OS that this deployment either (a) runs on (see URL column) or (b) has been tested on. Put in parens the OS number if more than OS instance is involved
3. Runtime - Runtime environment for component (e.g. JVM & version, Tomcat & version, etc.)
4. Assemble & Deploy
   • Links to documentation
5. Binding - how will externally consumed services of deployment scenario be consumed
6. Open - open enhancements and bugs (Bugzilla) for this deployment scenario (Note: none are currently defined)
7. URL - endpoint that hosts a test service (hosted by Eclipse Foundation)
8. Owner - person with overall responsibility for this deployment scenario (not individual components)

See Also

• Higgins Home