Metrics in Pomerium provide observability and monitoring data from your Pomerium deployment. Use metrics to review traffic and its effects on your system. Pomerium exposes standard Prometheus metrics related to its operation. Open-source users must [configure](https://www.pomerium.com/docs/reference/metrics.md) metrics collection and visualization manually. Pomerium Zero and Pomerium Enterprise provide simplified, built-in metrics collection and visualizations. ## Metrics definitions - **Traffic**: requests proxied by Pomerium to routes defined in a Pomerium deployment. - **Route**: the location of an upstream service protected behind Pomerium. At a minimum, a route consists of a [From](https://www.pomerium.com/docs/reference/routes/from.md) URL and a [To](https://www.pomerium.com/docs/reference/routes/to.md) URL. ## How metrics work Pomerium collects and aggregates traffic data from your Pomerium deployment. This data includes the: - request size, duration, and rate - number of authorized and denied requests - response code distribution You can view and filter this data in your traffic dashboard to gauge how much demand is placed on your server. ### Filter metrics #### Time range and routes Pomerium stores traffic metrics up to the last 30 days of usage. You can filter traffic by time range and routes. The following time ranges are supported: - Last 30 days - Last 2 weeks - Last 7 days - Last 24 hours - Last 12 hours - Last 3 hours - Last hour - Last 15 minutes (Enterprise only) When filtering by route, select: - **All Routes** to review aggregated traffic metrics across all routes defined in your Pomerium deployment. - an individual route to review aggregated traffic metrics for that route. **Zero:** \[Filtering metrics by time range routes in Pomerium Zero] **Enterprise:** \[Filtering metrics by time range routes in the Pomerium Enterprise console] ## Traffic dashboard **Zero:** Pomerium Zero collects traffic metrics at the [cluster](https://www.pomerium.com/docs/internals/clusters.md) level, which includes active replicas. To review the traffic dashboard in Pomerium Zero: 1. In the left-hand sidebar, select **Reports**. 2. Select **Traffic**. \[Find traffic metrics in Pomerium Zero] **Enterprise:** When you access the Enterprise Console, you'll land on the traffic dashboard. Pomerium Enterprise organizes traffic metrics in your deployment with [namespaces](https://www.pomerium.com/docs/internals/namespacing.md). Namespaces follow a hierarchical system. You can view traffic metrics for all namespaces, or a specific namespace, using the namespace dropdown menu. You can filter by: - Global namespace, which encompasses all namespaces in your deployment. - Parent namespace, which includes child namespaces (if any). - Child namespace, which displays metrics only for that namespace. \[Selecting a namespace to view its traffic in the Enterprise Console] In Pomerium Enterprise, you must configure metrics before you can view them. Metrics are not enabled by default, and are not required to run Pomerium Enterprise. See the [Configure Metrics](https://www.pomerium.com/docs/deploy/enterprise/configure-metrics.md) guide to enable metrics in your Enterprise deployment. ### Total and Authorized requests **Zero:** The **Total requests** chart shows the total number of proxied requests. The **Authorized requests** chart shows the total number of requests Pomerium authorized and forwarded to an upstream service. Both charts display the difference in requests between the selected and previous time ranges. \[Displaying the total and authorized requests in Pomerium Zero] The **Authorized Requests** pie chart displays the total number of authorized and denied requests. \[A pie chart showing the number of authorized and denied requests in Pomerium Zero] **Enterprise:** The **Total requests** chart shows the total number of proxied requests. The **Authorized requests** chart shows the total number of requests Pomerium authorized and forwarded to an upstream service. The **Healthy Endpoints** chart displays the number of healthy upstream endpoints, and roughly correlates with the number of routes defined in your deployment. For example, if a route's **To** definition includes [multiple upstream resources](https://www.pomerium.com/docs/reference/routes/to.md#target-multiple-upstream-resources), Pomerium includes these resources in the total sum of healthy endpoints. Pomerium excludes unhealthy endpoints from this total. See Load Balancing - [Active Health Checks](https://www.pomerium.com/docs/capabilities/routing.md#active-health-checks) and [Passive Health Checks](https://www.pomerium.com/docs/capabilities/routing.md#passive-health-checks) for more information. \[Viewing the total and authorized request charts in the Enterprise Console] The **Authorized Requests** pie chart displays the total number of authorized and denied requests. \[A pie chart showing the number of authorized and denied requests in Pomerium Enterprise] The **Healthy Upstream Endpoints** graph shows you the number of healthy endpoints over time. A dip in the graph denotes unhealthy endpoints. \[A graph displaying the number of healthy upstream endpoints in Pomerium Enterprise] ### Request durations **Zero:** Request duration measures the amount of time it takes Pomerium to proxy a request in milliseconds (ms). Pomerium Zero provides two request duration charts: The first chart organizes requests by duration ranges defined along the x-axis. Pomerium sums the total value of requests within each range and calculates the amount as a percentage value. \[A chart displaying request duration in Pomerium Zero] The second chart organizes requests by percentile ranges, date, and time. \[A chart displaying request duration in Pomerium Zero] **Enterprise:** Request duration measures the amount of time it takes Pomerium to proxy a request in milliseconds (ms). Pomerium Enterprise provides two request duration charts: The first chart organizes requests by duration ranges defined along the x-axis. Pomerium sums the total value of requests within each range and calculates the amount as a percentage value. \[A chart displaying request duration in Pomerium Enterprise] The second chart organizes requests by percentile ranges, date, and time. You can select a specific percentile range to view more granular data in that range: \[A chart displaying request duration with percentile ranges in Enterprise Console] ### Requests rate and response codes **Zero:** The **Requests per second** chart calculates the average amount of proxied requests per second over the span of an hour. Requests are organized by date and time, and categorized by the following response status codes: - **200s** (200-299): successful responses - **300s** (300-399): redirection messages - **400s** (400-499): client error responses - **500s** (500-599): server error responses \[A chart displaying requests per second in Pomerium Zero] **Enterprise:** The **Request Rate** chart calculates the average amount of proxied requests per second over the span of an hour. \[A chart displaying requests per second in Pomerium Enterprise] The **Response Codes** chart organizes requests by date and time, and categorizes them with the following response status codes: - **HTTP 2xx** (200-299): successful responses - **HTTP 3xx** (300-399): redirection messages - **HTTP 4xx** (400-499): client error responses - **HTTP 5xx** (500-599): server error responses \[A response codes chart in Pomerium Enterprise] ### Bytes sent and received, request size **Zero:** The **Bytes sent** and **Bytes received** charts display the average amount of bytes sent and received over the span of an hour. \[A chart displaying bytes sent and received in Pomerium Zero] **Enterprise:** The **Request Size** chart organizes requests based on their request size measured in kilobytes (KB). \[A chart displaying response size in Pomerium Enterprise] ## Runtime dashboard **Zero:** Runtime metrics are not supported in Pomerium Zero. **Enterprise:** In the **Runtime** dashboard, you can monitor how many system resources Pomerium is consuming. Filter by date range, service, and instance. \[The Runtime Info page in Pomerium Enterprise]