Crash Handling of Backend Servers

There are 3 different servers to execute a specific executable such as ICC, Router and Activation Manager (AM). Request from ICC goes to Router and get forwarded to Activation Manager (AM). Then AM executes the task and sends result back to Router which further gets forwarded to ICC.

If any of the server crashes while executing the batch and when recovery happens, the status is sent back to ICC server.

1. Router goes down: When router goes down, the Task Status will become indeterminate and the Batch Status will become Failed.
2. AM goes down: If AM goes down while executing a task, as soon as AM comes up, status of all tasks in the Batch will change to Indeterminate and the Batch Status will become Failed.
3. ICC goes down: When ICC goes down, the status of the task will become interrupted and the Batch Status will become Failed.
   • ICC will mark all the task status as interrupted even though some of the tasks might have executed successfully.
   • You have to manually validate the data before you re-trigger the batch again.