Terms
This chapter explains how an Axon Ivy Designer Project can be deployed to an Axon Ivy Engine. Before deploying an Axon Ivy project it is important to understand some major concepts and terms of the Axon Ivy Engine. The following chapter introduces these concepts and terms.
Application
Application "Webshop" (Roles, Users, Databases)
├── ProcessModel "Shop"
│ └── ProcessModelVersion 1
│ └── ProcessModelVersion 2
└── ProcessModel "BackEnd"
└── ProcessModelVersion 1
On the Axon Ivy Engine, applications can be configured. An application spans up an environment in which process models exist.
Process Model
Application "Webshop"
├── ProcessModel "Shop" (Project "Shop")
│ └── ProcessModelVersion 1
│ └── ProcessModelVersion 2
└── ProcessModel "BackEnd"
└── ProcessModelVersion 1
Within an application multiple process models can be configured. A process model on the Engine corresponds to an Axon Ivy project on the Designer. The difference is that a process model may hold multiple different versions of the same Axon Ivy project. A process model version - as its name suggests - is a version of an Axon Ivy project. In fact this version represents the state of an Axon Ivy project at the time it was deployed on the Engine.
Process Model Version
Application "Webshop"
├── ProcessModel "Shop"
│ └── ProcessModelVersion 1 (Project "Shop", V1, 12 cases)
│ └── ProcessModelVersion 2 (Project "Shop", V2, 304 cases)
└── ProcessModel "BackEnd"
└── ProcessModelVersion 1
A process model can have multiple versions called process model versions. These versions allow to change an Axon Ivy project without worrying about the compatibility of currently running cases on the Engine. How does this work? When an Axon Ivy project has been finished or reached a milestone, it is going to be deployed as the first process model version. Users can use this project, they start processes. Some of the processes may last long time (weeks, months, or even years). While these processes (e.g. cases) are running, the project may be enhanced and might have now incompatible changes to the first version. Now the changed project can not be deployed to the first version but to a new configured version. Consequently old cases must not be stopped, they will be still executed within the first process model version. Meanwhile new cases are started from the new deployed version.
A process model version has a release state. The release state of a process model version is responsible how the version is used by the system. The most important release state is the state RELEASED. Within a process model only one version can be in this state. All processes that are started in a process model are started in the released process model version! A complete list of release state can be found in the following list:
Name |
Description |
---|---|
PREPARED |
The process model version has been created and the project may already have been deployed. However, the process model version is not yet used. |
RELEASED |
The process model version is the currently released version. This means that all new processes are started in this version. Program Start and Web Service Process are only active for process model versions in this state. |
DEPRECATED |
The process model version has previously been in state RELEASED, but then another version was released. Therefore, this version is now not in RELEASED state but in DEPRECATED state. All cases that were started in this process model version will continue to run in this version. As soon as all cases of this version have been ended, the state will change to ARCHIVED automatically. |
ARCHIVED |
The process model version has previously been in state RELEASED, but then another version was released, and running cases has been finished in this process model. Consequently, this version is now not in RELEASED state anymore but has been ARCHIVED. Actually the engine administrator can change a process model version from state ARCHIVED back to state RELEASED if necessary. |
DELETED |
The process model version has been deleted. All project data belonging to this version has been deleted. |
The following diagram shows all release states and state changes that are possible:
Configuration Example
The following table shows an example of how applications, process models and process model versions on an Axon Ivy Engine can be configured.
Application |
Process Model |
Process Model Version |
Description |
---|---|---|---|
Company1 |
Application for company1. All users of the company are automatically imported to this application from the company’s active directory server. |
||
HRM |
Human Resource Management process model. Corresponds to the Axon Ivy project called “HRM”. |
||
V1 |
The first version of the HRM project was released in February 2008. This version is deprecated. There are still cases running in this version |
||
V2 |
The second version of the HRM project was released in April 2008. This version is released. All new processes are started in this version. |
||
V3 |
The third version of the HRM project was created in January 2009. This version is prepared, but not used productive. It will be released on the first of September 2009. |
||
Finance |
Finance process model. Corresponds to the Axon Ivy project Finance. |
||
V1 |
The first version of the Finance project was released in August 2007. This version is released. All new process are started in this version. |
||
Company2 |
Application for company2. The users of the company are managed by the Axon Ivy Engine. |
||
HRM |
Human Resource Management process model. Corresponds to the Axon Ivy project called “HRM”. |
||
V1 |
The first version of the HRM project was released in April 2008. This version is released, so that all HRM processes of company2 run and are started in this version. |