SevOne Data Platform REST API v3 SwaggerUI Guide & Request Rate Guidance

Overview

To find complete specifications of available REST API v3 endpoints, including requests, responses, and error codes, enabling you to effectively monitor, analyze, and manage your network's performance, please refer to IBM SevOne REST API v3 Hub > SevOne API v8.2.0 for details.

Whether you are looking to automate reports, configure alerts, or query performance metrics, the interactive API documentation will help you streamline your operations with ease. Explore the capabilities of IBM SevOne NPM APIs to unlock the full potential of your network performance management.

API Performance and Request Rate Recommendation

The SevOne v3 API is a highly flexible and powerful interface that enables users to perform a wide range of operations. However, this level of flexibility can also introduce risk if the API is subjected to excessive or unmanaged load.

As with any API, as request volume increases, each in-flight request consumes system resources, including memory for requested objects, parsed payloads, database results, and response serialization. At the same time, CPU utilization rises as the system processes a higher number of concurrent operations. When memory limits are approached, performance will be negatively impacted and in some extreme scenarios, the system may terminate processes (e.g., via the OOM killer), or the application itself may become unstable or crash.

Based on performance testing conducted on multiple clusters in our test environments, IBM recommends limiting API request rates to no more than 8,000 requests per minute per cluster to help ensure reliable and stable operation. Actual performance may vary depending on your deployment configuration, hardware resources, cluster size, and workload characteristics.

Important: Some API requests are heavier than others because they require more computational resources, process larger datasets, perform complex calculations, or involve multiple queries and external service calls, resulting in longer response times and higher server load.