Initialization on non-master nodes

[GlennRicaud]

With a clean cluster configured with 3 nodes (6.14.3):

• Start first node
  • The node waits or is in RED state
• Starts the second node
  • Multiple exceptions when starting:

    com.enonic.xp.repository.RepositoryNotFoundException: Repository with id [system-repo] not found
    at com.enonic.xp.repo.impl.node.NodeServiceImpl.lambda$verifyBranchExists$10(NodeServiceImpl.java:982)
    at com.enonic.xp.context.ContextImpl.runWith(ContextImpl.java:85)
    at com.enonic.xp.repo.impl.node.NodeHelper.runAsAdmin(NodeHelper.java:15)
    at com.enonic.xp.repo.impl.node.NodeServiceImpl.verifyBranchExists(NodeServiceImpl.java:976)
    at com.enonic.xp.repo.impl.node.NodeServiceImpl.verifyContext(NodeServiceImpl.java:958)
    at com.enonic.xp.repo.impl.node.NodeServiceImpl.refresh(NodeServiceImpl.java:711)
    at com.enonic.xp.core.impl.app.ApplicationRepoServiceImpl.getApplications(ApplicationRepoServiceImpl.java:90)
    at com.enonic.xp.core.impl.app.ApplicationServiceImpl.doInstallStoredApplications(ApplicationServiceImpl.java:201)

Visually, this has, for example, as a result that the IssueService is not started (->ContentServiceImpl->ContentTypeServiceImpl->ApplicationServiceImpl):

[GlennRicaud]

Also noticed that IssueService does not check that it is the master node (Not the problem here, but should be fixed)

[GlennRicaud]

=> Initialization of services on non-master nodes should be blocking until the end of the Initializer procedure on the master node. => Refactor the initializers to handle this case. Refactor to make the code common to all initializer