Metrics

Atlasflow records resource metrics for your running deployments so you can understand performance and right-size your workload.

# Available metrics

Metrics are displayed as time-series charts on the deployment view. You can change the time range to zoom in on a spike or look at a longer trend.

# Reading the charts

• A sustained high CPU line suggests your app is compute-bound; consider optimizing hot paths or running more instances (available on the Team plan).
• A memory line that climbs steadily and never drops can indicate a leak. A line that approaches the container's limit risks the process being terminated.

# Metrics and billing

The same resource usage that appears in these charts is what Atlasflow meters for usage-based billing. Watching metrics is a good way to understand and predict your bill.

# Next steps

• Usage and credits: how usage is metered and billed.
• Runtime logs: correlate metrics with log output.
• Build machines: resources for builds (separate from runtime).