File monitoring in HCL Universal Orchestrator

​With technology advancing at breakneck speed, real-time monitoring of modifications to a file can play a significant role in maintaining efficiency and reliability of modern applications. A feature that identifies changes to a file as they occur proves helpful in various aspects of software workflows and operations. 

HCL Universal Orchestrator introduces such a feature in this release: file monitoring that instantly detects and responds to file changes, streamlining automation, minimizing delays, and boosting productivity.
With file monitoring, users can:

• Detect errors in log files.

• Track configuration file changes.
• Identify incoming data.

This continuous oversight ensures that any changes are quickly identified, accelerating the process and ensuring seamless data processing.
The behaviour of file monitor tasks is largely influenced by the Polling type attribute, and its combination with multiple attributes used in different scenarios. The feature supports two polling types: eventDriven and periodic, each suited for specific use cases. You can monitor the files on a local machine or a mounted virtual machine where the HCL UnO agent is installed. It detects file events such as creation, modification, and deletion, ensuring timely recognition and processing of critical changes.
The task runs continuously until either a change is detected, or the predefined monitoring period expires. The results, including affected file names and event types, are logged in the output for further analysis.
How to Create a File Monitoring Task

1. Log in to the HCL UnO UI.
2. Open the Design.
3. Add the File Monitoring Task template under File Transfer and Coordination.

Configuration Details
            All the available attributes for file monitoring tasks are categorized under two tabs.
General information

• Name: specify a mandatory name for the task.
• Workstation: define the workstation (must be a standalone agent).

Action

1. Event: choose the event type (Create, Modify, or Delete).
2. File path: provide the file path to monitor a single file or multiple files.
3. Minimum file size: set the minimum file size for monitoring (in bytes).
4. Monitoring time: duration (in seconds) to monitor file changes.
5. Post modification time: monitors the file for additional time after detecting changes (Create or Modify only).
6. Exit if the path to the monitoring file is not found: task moves to ABEND status, if the file path is invalid.
7. Output file path: specify where you want the output file after the task completion.
8. Polling Type: define the type of monitoring for the file:
   • eventDriven: the task moves to SUCC status immediately after detecting a change or when the monitoring period elapses.
   • Periodic: the file is monitored at specified intervals.

For more information, see Analyzing file monitoring task scenarios.

1. Record all modified files (Batch): enable this option to log all modified file names by setting the value to "". If not, only the first modified file is logged.

Analysing file monitoring task scenarios
You can configure the polling type attributes in multiple ways to monitor the files for specific use cases.

• Event driven polling
  •  First Run: Regardless of the batch attribute, the task captures the first detected change and moves to SUCC status.
  •  Rerun Behaviour: If files change before a rerun, the task compares the stored file state with the current state. It waits until the time specified for Post modification time and moves to SUCC immediately otherwise.  If you enable the Batch attribute before rerun, the task records all detected changes before the rerun. This is the only scenario where multiple file changes are recorded for eventDriven polling type. If not, only the first changed file is recorded.
• Periodic Polling
  • Initial and Rerun Behaviour: The task periodically checks for file changes. If any file changes before the rerun, the task behaves like an event-driven task and moves to a SUCC status. You can enable the Batch attribute to record all the changed files in the output log. If not, only the first changed file (in alphabetical order) is recorded.

Author's Bio

---

​Shubham Gupta (Senior Software Engineer)
Working for the Workload Automation Bangalore team as a backend developer. He worked with platforms including Java, Spring, Spring boot, Spring MVC, SQL etc. He enjoys practicing yoga and derives great joy from exploring new places and getting to know new people.