You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 3 Next »

Introduction

  • An outage of JOC Cockpit in most situations does not affect execution of workflows and jobs. Due to a resilient architecture the JS7 components work autonomously.
  • JOC Cockpit is used to manage the inventory of objects and to monitor the execution of workflows by Controllers and Agents. This means that in case of an outage you are blind, but it doesn't mean that jobs are not running.
  • If JOC Cockpit is not available then 
    • users cannot login to the JOC Cockpit and cannot check the status of workflow execution by orders. Users cannot check the logs of workflows and jobs.
    • users cannot manage the inventory, modify and deploy workflows and related objects. No orders can be submitted.

  • Testing by SOS includes to performed successful tests for the scenario when the JOC Cockpit is not available for 24 hours and the Agent executes all scheduled orders. When the JOC Cockpit is started again then job execution results in the History are updated and become visible with JOC Cockpit.

Controller Behavior

  • The Controller forwards execution results of workflows to JOC Cockpit. This information is stored with the Controller's journal. If no connection to JOC Cockpit is available then
    • the Controller's journal will grow. Depending on the load of workflows the journal files that are stored with the ./state directory, can grow to some Gigabytes. There is no harm about this as long as sufficient disk space is available.
    • this does not affect the communication between Controller and Agents. Agents will report back order state transitions and execution results to the Controller.
  • When the JOC Cockpit outage is resolved then JOC Cockpit will re-establish the connection to the Controller. At this point in time the information about execution results will become available for the JOC Cockpit History Service that stores the information with the database.
    • Depending on the load of workflows and the duration of the outage it can take some time to update the JOC Cockpit History. However, no information is lost.
    • With execution results being added to the JOC Cockpit database the Controller's journal will shrink.

Agent Behavior

  • Workflows are deployed to Agents just once. When the Agent receives orders for workflows some time in advance, e.g. by the Daily Plan, then it can execute them autonomously.
  • Agents will continue to execute workflows for the scheduled time. If the connection between Controller and Agent is available then Agents will report back to the Controller.



  • No labels