org.camunda.bpm.engine.repository

Interface DeploymentHandler

All Known Implementing Classes:

DefaultDeploymentHandler

public interface DeploymentHandler

The DeploymentHandler interface should be implemented when there is a need to define a custom behavior for determining what Resources should be added to a new Deployment.

The class that implements this interface must provide a Process Engine parameter in it's constructor. Custom implementations of this interface should be coupled with an implementation of the DeploymentHandlerFactory interface, which can then be wired to the Process Engine through the ProcessEngineConfigurationImpl.setDeploymentHandlerFactory(DeploymentHandlerFactory) method.

See DefaultDeploymentHandler for the default behavior of the deployment process.

Method Summary

Modifier and Type	Method and Description
Set<String>	determineDeploymentsToResumeByDeploymentName(CandidateDeployment candidateDeployment) This method is called in the last stage of the deployment process, if a Process Application deployment is performed.
Set<String>	determineDeploymentsToResumeByProcessDefinitionKey(String[] processDefinitionKeys) This method is called in the last stage of the deployment process, if a Process Application deployment is performed.
String	determineDuplicateDeployment(CandidateDeployment candidateDeployment) This method is called in the second stage of the deployment process, when the previously called shouldDeployResource(Resource, Resource) method determines that none of the provided resources for deployment satisfy the comparison criteria, i.e.
boolean	shouldDeployResource(Resource newResource, Resource existingResource) This method is called in the first stage of the deployment process.

Method Detail

shouldDeployResource

boolean shouldDeployResource(Resource newResource,
                             Resource existingResource)

This method is called in the first stage of the deployment process. It determines if there is a difference between a Resource that is included in the set of resources provided for deployment, and a Resource of the same name, already deployed to the Process Engine with a previous Deployment. The method will be called for every (new, existing) Resource pair. If a Resource of the same name doesn't exist in the Process Engine database, the new Resource is deployed by default and this method is not called.

An implementation should define the comparison criteria for the two provided Resources. For the default comparison criteria, see DefaultDeploymentHandler.shouldDeployResource(Resource, Resource).

The output of this method (Boolean) will determine if newResource is included in the list of Resources to be deployed (true), or not (false).

Parameters:

newResource - is a resource that is part of the new Deployment.

existingResource - is the most recently deployed resource that has the same resource name.

Returns:

true, if the new Resource satisfies the comparison criteria, or false if not.

determineDuplicateDeployment

String determineDuplicateDeployment(CandidateDeployment candidateDeployment)

This method is called in the second stage of the deployment process, when the previously called shouldDeployResource(Resource, Resource) method determines that none of the provided resources for deployment satisfy the comparison criteria, i.e. there are no new Resources to deploy. This method then determines a Deployment already present in the Process Engine database that is a duplicate of the new Deployment.

An implementation should determine what defines a duplicate Deployment.

The returned Deployment ID (String) will be used to retrieve the corresponding Deployment from the Process Engine database. This Deployment will then be registered with the JobExecutor and the Process Application (if a Process Application Deployment is performed). Furthermore, any Process Definitions that this Deployment contains will be scheduled for activation, if a Process Definition Activation Date was set.

Parameters:

candidateDeployment - a wrapper for the set of Resources provided for deployment and the name under which they should have been deployed. At this stage, it should be assumed that all of the Resources of this set didn't satisfy the comparison criteria and will not be deployed.

Returns:

the Deployment ID of the Deployment determined to be a duplicate of the data provided by the candidateDeployment.

determineDeploymentsToResumeByProcessDefinitionKey

Set<String> determineDeploymentsToResumeByProcessDefinitionKey(String[] processDefinitionKeys)

This method is called in the last stage of the deployment process, if a Process Application deployment is performed. An additional condition is that the ProcessApplicationDeploymentBuilder.resumePreviousVersions() flag of the Process Application Deployment Builder is set, and ProcessApplicationDeploymentBuilder.resumePreviousVersionsBy(String) ()} is set to the default, ResumePreviousBy.RESUME_BY_PROCESS_DEFINITION_KEY value.

The implementation of this method determines a Set of Deployment IDs of Deployments already present in the Process Engine database. These deployments will then be registered with the Process Application, so that any Java classes that the Process Application provides can be utilised by the Process Definitions of the Deployments.

Parameters:

processDefinitionKeys - are the Process Definition Keys of a subset of Process Resources provided for deployment. The subset includes Process Definition Keys from:

• Processes from Resources that will not be deployed to the Process Engine database due to the outcome of the shouldDeployResource(Resource, Resource) method.
• Processes from Resources that will be deployed to the Process Engine database, that update an existing Process Definition, i.e. there is a Process Definition with the same key present in the Process Engine database.

It should be noted that, if shouldDeployResource(Resource, Resource) determines that all of the provided resources should be deployed to the database, and there are no Process Definitions with the same key already present in the database (i.e. the Process Resources are completely new), this method will not be called.

Returns:

a Set of deployment IDs of Deployments already present in the Process Engine database, that should be resumed (registered with the JobExecutor and registered with the newly deployed Process Application).

determineDeploymentsToResumeByDeploymentName

Set<String> determineDeploymentsToResumeByDeploymentName(CandidateDeployment candidateDeployment)

This method is called in the last stage of the deployment process, if a Process Application deployment is performed. An additional condition is that the ProcessApplicationDeploymentBuilder.resumePreviousVersions() flag of the Process Application Deployment Builder is set, and ProcessApplicationDeploymentBuilder.resumePreviousVersionsBy(String) ()} is set to the ResumePreviousBy.RESUME_BY_DEPLOYMENT_NAME value.

The implementation of this method should determine a set of Deployment IDs of Deployments with the same name, that are already present in the Process Engine database.

Parameters:

candidateDeployment - a wrapper for the set of resources that were provided for deployment and the name that should be used for this deployment. This information can be used to find any Deployments of the same name already present in the Process Engine database. The set of Resources can then be used, if needed, to filter out the resulting Deployments according to a given criteria.

Returns:

a Set of deployment IDs of Deployments already present in the Process Engine database, that should be resumed (registered with the JobExecutor and registered with the newly deployed Process Application).