Learning center"Four features of the Assistant API you aren't using - but should"Learn more
Preview Mode ()

Our updated Datadog integration and new Prometheus endpoints make discovering insights into your indexes' performance and usage easier than ever. Building upon existing health metrics, you can now also monitor request frequency and duration, consumption of read and write units, and core metrics on latency through our Datadog integration or directly via our new Prometheus endpoints.

Our Datadog integration now delivers metrics for both serverless and pod-based indexes, with 25 new metrics for more granular performance monitoring and improved observability. With the updated integration, you can:

  • Monitor the number of requests made and request duration by endpoint (e.g. the count of upsert requests, query requests, update requests, and more).
  • Explore and visualize your read and write unit consumption, plus the size of your serverless indexes.
  • Track usage patterns over time and easily identify anomalies with out-of-the-box dashboards. Customize your experience further by setting conditional formatting for selected values, viewing certain metrics side by side, or changing data visualizations.
  • Get alerted when the number of writes to your serverless index exceeds a specified threshold to avoid latency issues. This recommended monitor can be customized to meet your team’s specific configuration.
Our out-of-the-box Datadog dashboard for serverless surfaces health metrics on throughput, latency, read and write units, and more.

Getting started

To set up the Datadog integration, go to Datadog's Pinecone integration page. In the Configure tab, simply add your project ID and API Key for the project you want to monitor. Under Monitoring Resources, you'll find the dashboard and recommended monitors, with a full list of available metrics in the Data Collected tab.

For non-Datadog users, these new metrics are fully accessible via Prometheus. The endpoints are designed with HTTP service discovery to automatically detect and target all serverless indexes across regions within a project, making it easy to monitor the health and performance of your entire environment — no matter how it’s distributed.

To set up monitoring directly with Prometheus, you'll need your Pinecone API Key and project ID. Update the scrape_configs section of your prometheus.yml file with this information to start querying. For step-by-step instructions, a list of available metrics, and example queries, check out our Monitor with Prometheus guide.

Prometheus metrics and our Datadog integration are available to all users on Standard and Enterprise plans. Visit our documentation to learn more and start up-leveling your observability today.

Share:

What will you build?

Upgrade your search or chatbots applications with just a few lines of code.