ESP Workload Automation FILE_TRIGGERs when virtual agent has Active real agents.

Products

• ESP Workload Automation

Issue/Introduction

Have one virtual agent configured with two Active/Active real agents. How to make sure the FILE_TRIGGER (created on one real agent A) will continue monitoring after real agent A is stopped and no duplicate fires after real agent A is restarted? 

Note: The FILE_TRIGGER has CONTINOUS on its FILENAME.

Environment

Component: ESP Workload Automation

Release : 12.0

Resolution

Try avoiding this setup as it's very hard and complex to handle the FILE_TRIGGER wobs, especially for CONTINOUS file_triggers.

Concerns:
- Force completing the FILE_TRIGGER will send a FILETRIG DELETE request to the real agent. However if the agent is down, it can't be processed right away, and after the agent is started, this request may not be processed BEFORE the file detected notification. This can cause duplicate fires of same file activity.
- RESUB the failed FILE_TRIGGER won't generate the FILETRIG DELETE request, and will only send the FILETRIG DEFINE request to the currently active real agent.
- Every FILETRIG DEFINE will create a new thread with unique ID string. However ESP doesn't match it when processing received file detected notifications from the agent. This is the main reason for duplication.