Internet Transit - CoreWeave Docs

The Internet Transit dashboard provides a near-real-time view into how much traffic flows through your backbone egress and ingress points, how that traffic is distributed across internal services, and how well the network is performing from an end-user perspective. Use this dashboard to monitor Node and service-level traffic and connection statistics so you can detect incidents, plan capacity, and troubleshoot network issues. To view the dashboard, go to the Internet Transit dashboard.

For instructions about accessing CoreWeave Grafana dashboards, see Access and use CoreWeave Grafana dashboards.

You can use this dashboard to:

Spot spikes or drops in throughput that may indicate an incident or a saturation event.
Correlate traffic behavior with latency (RTT) and connection counts to understand user impact.
Drill into per-Node or per-service views for focused troubleshooting.
Identify which Nodes or services carry the most load so you can plan capacity.
Monitor how much traffic flows to and from third-party clouds.

Filters and parameters

Use these filters at the top-left of the page to choose the data you want to view:

Data Source: The Prometheus data source selector.
Zone: The to view.
Cluster Org: The organization that owns the cluster.
Cluster: The specific cluster in the organization to view.

Set the time range and refresh rate parameters at the top-right of the page. The default time range is 5 minutes, and the default refresh rate is 1 minute.

Panel descriptions

The dashboard has several panels with views into the network traffic and connection statistics.

Panel	Description
Traffic - Inbound Per Node	A sortable table where each row represents a Node. The Throughput column is rendered as a horizontal gauge, making it easy to see relative utilization at a glance. Click a Node to open the Node Details dashboard for deep-dive analysis.
Traffic - Outbound Per Node	The complement to Inbound Per Node, but for egress traffic. This panel is useful for determining which Nodes may require additional bandwidth or rate-limiting.
Traffic - Inbound Total	A single card showing the aggregate inbound rate (Gb/s).
Traffic - Outbound Total	The complement to Inbound Total, showing the aggregate outbound rate (Gb/s).
Average RTT per Service	A time-series graph showing the rolling 5-minute average round-trip time (RTT) for both internal and external paths, grouped by service signature. This panel helps identify latency trends and potential congestion points.
Traffic - Inbound Per Service	A stacked-area time-series chart that breaks down inbound traffic by service and categorized by third-party cloud vendors. The legend is interactive, letting you toggle visibility and sort by current throughput, minimum, or maximum. Note: New traffic is labeled (Being analyzed) until it’s processed and categorized.
Traffic - Outbound Per Service	The complement to Inbound Per Service, but for egress traffic. This panel is useful for determining which services are generating the most outbound traffic.
Traffic - Total	A time-series graph showing the total inbound and outbound traffic. Inbound is shown in green above the X-axis, and outbound is shown in yellow below. Helps visualize the net traffic balance and identify asymmetric patterns.
Connections - Total	A time-series graph showing the total number of concurrent connections across all hosts. Helps identify connection flood or drop-off events.
Connections - Per Node	A stacked-area time-series chart that breaks down the number of concurrent connections by Node. This helps to identify connection hot spots and potential load-balancing issues.
Connections - Per Service	A stacked-area time-series chart that breaks down the number of concurrent connections by service object. This helps to identify which services are holding the most active connections.

​Filters and parameters

​Panel descriptions

Filters and parameters

Panel descriptions