Table of Contents

Search

  1. Preface
  2. Workflows
  3. Workflow Variables
  4. Workflow Parameters
  5. Cluster Tasks
  6. Command Task
  7. Human Task
  8. Mapping Task
  9. Notification Task
  10. Gateways
  11. Workflow Recovery
  12. Workflow Administration

Developer Workflow Guide

Developer Workflow Guide

Workflow Recovery on Grid

Workflow Recovery on Grid

The status of a workflow that fails over from one node to another on a grid depends on the type of task that is running when the node fails.
A running workflow tries to continue to run during failover regardless of the workflow recovery strategy that you specify. If the currently running task is a Command task or a Mapping task at the point of failover, the workflow recovery strategy determines the workflow behavior.
If you specify an automatic recovery strategy for the workflow, the workflow reruns the Mapping task or the Command task from the start of the task. If you do not specify an automatic recovery strategy, the workflow is canceled and you must manually recover the workflow on the active node.
You can configure automatic recovery for a workflow that runs on a Data Integration Service grid. If the master service process shuts down unexpectedly, the master role fails over to another service process. The current master can automatically recover any workflow instance that was running during the failover and that you configured for automatic recovery.
Do not configure a workflow for automatic recovery if the workflow includes a Human task. If the workflow is interrupted during a Mapping task that precedes a Human task, the recovered workflow reruns the Mapping task and adds conflicting data to the target database.

0 COMMENTS

We’d like to hear from you!