--- title: Pinecone description: Cloud based Vector Database for high-performance AI applications. breadcrumbs: Docs > Integrations > Pinecone --- # Pinecone Integration version1.2.0 {% callout %} # Important note for users on the following Datadog sites: us2.ddog-gov.com {% alert level="info" %} To find out if this integration is available in your organization, see your [Datadog Integrations](https://app.datadoghq.com/integrations) page or ask your organization administrator. To initiate an exception request to enable this integration for your organization, email [support@ddog-gov.com](mailto:support@ddog-gov.com). {% /alert %} {% /callout %} Pinecone Pod-Based Dashboard OverviewPinecone Serverless Dashboard Overview ## Overview{% #overview %} {% alert level="info" %} **Note:** Pinecone has renamed several serverless metrics. Until **June 30, 2026**, both the legacy (`*_total`) and new (`*_count`/`*_duration_sum`) names are emitted by Pinecone, so existing dashboards and monitors continue to work during the transition. After that date, only the new names will be available. See Serverless metric name mapping for the full list. {% /alert %} - **Optimize performance and control usage**: Observe and track specific actions (e.g. request count) within Pinecone to identify application requests with high latency or usage. Monitor trends and gain actionable insights to improve resource utilization and reduce spend. - **Automatically alert on metrics**: Get alerted when index fullness reaches a certain threshold. You can also create your own customized monitors to alert on specific metrics and thresholds. - **Locate and triage unexpected spikes in usage or latency**: Quickly visualize anomalies in usage or latency in Pinecone's Datadog dashboard. View metrics over time to better understand trends and determine the severity of a spike. ## Requirements{% #requirements %} Monitoring Pinecone with Datadog requires: - An Enterprise or Enterprise Dedicated Pinecone plan. - Pod-based or Serverless indexes: Datadog supports both pod-based and serverless metric capturing. ## Setup{% #setup %} ### Installation{% #installation %} 1. Login to your [Pinecone account](https://app.pinecone.io/). 1. Navigate to **API Keys** tab. 1. Create an API key. 1. Copy the created API Key to your clipboard. ### Configuration{% #configuration %} 1. Navigate to the configuration tab inside Datadog [Pinecone integration tile](https://app.datadoghq.com/account/settings#integrations/pinecone). 1. Enter your Pinecone Project ID which can be found in the project list in the Pinecone console. 1. For pod-based environments only: Select your environment. Projects in serverless environments can leave this blank. 1. Paste your copied API key. ## Data Collected{% #data-collected %} ### Metrics{% #metrics %} | **Type** | **Description** | **Metric prefixes collected** | | ----------------- | ------------------------------------------------------------------------- | ----------------------------- | | **Account Usage** | Collection of metrics for determining number of records per pod in index. | `pinecone.vector` | | **Index** | Index storage utilization metrics, such as fullness. | `pinecone.index` | | **Latency** | Collection of server-side latency metrics for Pinecone data plane calls. | `pinecone.request` | | **Serverless** | Collection of metrics for indexes described as being type `Serverless`. | `pinecone.db` | | | | | | **pinecone.vector.count**(gauge) | Number of records per pod in the index.*Shown as record* | | **pinecone.request.count.total**(count) | Number of data plane calls made by clients.*Shown as request* | | **pinecone.request.error.count.total**(count) | Number of data plane calls made by clients that resulted in errors.*Shown as request* | | **pinecone.request.latency.seconds.min**(gauge) | Minimum of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.max**(gauge) | Maximum of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.avg**(gauge) | Average of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.50percentile**(gauge) | p50 of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.90percentile**(gauge) | p90 of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.95percentile**(gauge) | p95 of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.99percentile**(gauge) | p99 of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.99.9percentile**(gauge) | p99.9 of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as second* | | **pinecone.request.latency.seconds.count**(count) | Count of the distribution of server-side processing latency for Pinecone data plane calls.*Shown as request* | | **pinecone.index.fullness**(gauge) | Fullness of the index on a scale of 0 to 1.*Shown as unit* | | **pinecone.db.op.query.total**(count) | The number of Query Request made to an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.query.count.*Shown as request* | | **pinecone.db.op.fetch.total**(count) | The number of Fetch Request made to an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.fetch.count.*Shown as request* | | **pinecone.db.op.update.total**(count) | The number of Update Request made to an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.update.count.*Shown as request* | | **pinecone.db.op.delete.total**(count) | The number of Delete Request made to an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.delete.count.*Shown as request* | | **pinecone.db.op.upsert.total**(count) | The number of Upsert Request made to an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.upsert.count.*Shown as request* | | **pinecone.db.op.query.duration.total**(count) | Total time taken processing Query Request for an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.query.duration.sum.*Shown as millisecond* | | **pinecone.db.op.fetch.duration.total**(count) | Total time taken processing Fetch Request for an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.fetch.duration.sum.*Shown as millisecond* | | **pinecone.db.op.update.duration.total**(count) | Total time taken processing Update Request for an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.update.duration.sum.*Shown as millisecond* | | **pinecone.db.op.delete.duration.total**(count) | Total time taken processing Delete Request for an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.delete.duration.sum.*Shown as millisecond* | | **pinecone.db.op.upsert.duration.total**(count) | Total time taken processing Upsert Request for an Index (Serverless). Deprecated 2026-06-30; use pinecone.db.op.upsert.duration.sum.*Shown as millisecond* | | **pinecone.db.write.unit.total**(count) | Total number of write units consumed (Serverless). Deprecated 2026-06-30; use pinecone.db.write.unit.count.*Shown as request* | | **pinecone.db.read.unit.total**(count) | Total number of read units consumed (Serverless). Deprecated 2026-06-30; use pinecone.db.read.unit.count.*Shown as request* | | **pinecone.db.op.query.count**(count) | The number of Query Request made to an Index (Serverless).*Shown as request* | | **pinecone.db.op.fetch.count**(count) | The number of Fetch Request made to an Index (Serverless).*Shown as request* | | **pinecone.db.op.update.count**(count) | The number of Update Request made to an Index (Serverless).*Shown as request* | | **pinecone.db.op.delete.count**(count) | The number of Delete Request made to an Index (Serverless).*Shown as request* | | **pinecone.db.op.upsert.count**(count) | The number of Upsert Request made to an Index (Serverless).*Shown as request* | | **pinecone.db.op.query.duration.sum**(count) | Total time taken processing Query Request for an Index (Serverless).*Shown as millisecond* | | **pinecone.db.op.fetch.duration.sum**(count) | Total time taken processing Fetch Request for an Index (Serverless).*Shown as millisecond* | | **pinecone.db.op.update.duration.sum**(count) | Total time taken processing Update Request for an Index (Serverless).*Shown as millisecond* | | **pinecone.db.op.delete.duration.sum**(count) | Total time taken processing Delete Request for an Index (Serverless).*Shown as millisecond* | | **pinecone.db.op.upsert.duration.sum**(count) | Total time taken processing Upsert Request for an Index (Serverless).*Shown as millisecond* | | **pinecone.db.write.unit.count**(count) | Total number of write units consumed (Serverless).*Shown as request* | | **pinecone.db.read.unit.count**(count) | Total number of read units consumed (Serverless).*Shown as request* | | **pinecone.db.storage.size.bytes**(gauge) | Total size of the index in bytes (Serverless)*Shown as byte* | | **pinecone.db.record.total**(gauge) | Total number of records (Serverless)*Shown as record* | ### Logs{% #logs %} Pinecone does not include logs collection. ### Service Checks{% #service-checks %} Pinecone does not include any service checks. ### Events{% #events %} Pinecone does not include any events. ## Troubleshooting{% #troubleshooting %} ### Serverless metric name mapping{% #serverless-metric-name-mapping %} Pinecone has renamed several serverless metrics. Until **June 30, 2026**, both the legacy and new metric names are emitted, so existing dashboards and monitors continue to work. After that date, only the new names will be available — update any saved queries to use the new metric names below. | Legacy metric (removed 2026-06-30) | New metric | | -------------------------------------- | ------------------------------------ | | `pinecone.db.op.upsert.total` | `pinecone.db.op.upsert.count` | | `pinecone.db.op.upsert.duration.total` | `pinecone.db.op.upsert.duration.sum` | | `pinecone.db.op.query.total` | `pinecone.db.op.query.count` | | `pinecone.db.op.query.duration.total` | `pinecone.db.op.query.duration.sum` | | `pinecone.db.op.fetch.total` | `pinecone.db.op.fetch.count` | | `pinecone.db.op.fetch.duration.total` | `pinecone.db.op.fetch.duration.sum` | | `pinecone.db.op.update.total` | `pinecone.db.op.update.count` | | `pinecone.db.op.update.duration.total` | `pinecone.db.op.update.duration.sum` | | `pinecone.db.op.delete.total` | `pinecone.db.op.delete.count` | | `pinecone.db.op.delete.duration.total` | `pinecone.db.op.delete.duration.sum` | | `pinecone.db.write.unit.total` | `pinecone.db.write.unit.count` | | `pinecone.db.read.unit.total` | `pinecone.db.read.unit.count` | ## Support{% #support %} Need help? Contact [Datadog support](https://docs.datadoghq.com/help/). ## Further Reading{% #further-reading %} - [Integration roundup: Monitoring your AI stack](https://www.datadoghq.com/blog/ai-integrations/) - [Pinecone](https://docs.datadoghq.com/integrations/pinecone.md)