Optimizing ServiceNow: Achieving Scale with Multiple Nodes

[Omkar Mone]

I’m excited to share a recent achievement in leveraging ServiceNow’s powerful architecture to significantly enhance performance and scalability.

Challenge:

We needed to create ~1 million records with multiple related records to be created with attachments with ~4mb as well in ServiceNow. It takes approx. ~7-8 days to create this amount of data in ServiceNow if only single node is utilised. This is a task that demands efficiency and speed to minimize system load and ensure smooth operations.

If we would have run a script on a single node, the script would run in the scheduler taking precedence over other activities which would slow down the instance and impact the performance.

Solution: Utilizing Scheduled Jobs, Events, and Script Actions:

To tackle this, we developed a strategy that harnesses the power of ServiceNow’s nodes (we had 20 primary nodes). By using scheduled jobs tied to specific nodes to fire events and script actions, we were able to distribute the workload.

Here's a breakdown of our approach:

1. Scheduled Jobs:

   - We set up scheduled jobs to initiate the record creation process.

   - These jobs were configured to run on specific nodes, ensuring an even distribution of the workload across the environment.

2. Event Firing:

   - Each scheduled job was designed to fire a custom event upon execution.

   - This event acted as a trigger for the record creation script action.

3. Script Action:

   - A robust script action was associated with the custom event.

   - This script was responsible for the actual creation of records, including error handling and logging mechanisms to ensure data integrity and facilitate troubleshooting.

4. Node-Specific Execution:

   - By tying scheduled jobs to specific nodes, we achieved parallel processing.

   - Each node handled a portion of the record creation task, significantly reducing the overall time required and preventing any single node from becoming a bottleneck.

Interestingly, by this, we were also able to have a dedicated node, running all its 8 semaphores dedicated to doing one job, which significantly enhanced the performance.

I cannot imagine many other gates that this would open.

Outcome:

The result was remarkable! We successfully created ~1 million records within 2 days than it would have taken with a single-node approach. This showcased the scalability of ServiceNow and highlighted our team’s ability to innovate and optimize processes, which gained a lot of trust from the customers in the platform.

 Impact:

- Efficiency: Drastically reduced the time required for bulk record creation.

- Scalability: Demonstrated ServiceNow’s capability to handle large-scale operations.

- Performance:  Maintained system performance and stability throughout the process.

Feel free to reach out to me in case of anything.

Regards,

Omkar Mone