Logs for a Magic Containers application live under Logging in the left-side menu of the application. The Logs page shows what is happening in your application right now and what happened over the past few days. The Settings page is where you configure log forwarding.
The Logs page is split into three tabs:
All three tabs stream new entries live. Use the pause button next to the search box to freeze the view while you read, and the date picker to look back at earlier entries. Click any row to expand it and see more detail about that entry.
Application logs
The Application tab shows everything your containers write to stdout and stderr. If your application prints a log level along with the message, the level shows up as its own column so errors and warnings stand out from the rest.
Logs are kept for up to 5 days. The date picker sets the range you want to look at, and the live stream keeps adding new rows at the bottom while the range includes today.
You can narrow the view with the following filters:
- Container to see the output of a single container in a pod.
- Pod to focus on one pod.
- Region to check whether a problem is limited to one location.
- Keyword search to find a specific message.
Narrowing down to a single region is a quick way to spot regional issues, like network latency or a hardware constraint in one location. If the same error shows up across pods and regions at the same time, the cause is more likely in the application itself.
System logs
The System tab shows the events the platform generates while running your application. This includes deployments, image pulls, container starts and stops, scaling, and networking setup. These logs are the place to look when a deployment does not go as expected or a container keeps restarting.
System logs work the same way as application logs. They are kept for up to 5 days, each row expands to show more detail, and the same filters apply: container, pod, region, and keyword search.
Switching between the Application and System tabs is often the fastest way to tell whether a problem comes from the platform or from your code. If the System tab shows a healthy deployment and the Application tab shows errors, the issue is in the application. If the System tab shows a failed start or a health check problem, start there.
HTTP logs
The HTTP tab appears when your application has CDN endpoints. It shows the requests those endpoints served, with the status code, cache status, edge location, client IP, path, and response size for each one.
HTTP logs are kept for up to 3 days. Each row expands to show more detail about the request.
HTTP logs have their own set of filters:
- Pull zone
- HTTP status code
- Cache status
- Country
- Edge location
- URL
- User agent
- Keyword search
Filtering by status code is a good starting point when users report errors. Pair it with the cache status to see whether the failing requests reached your containers or were answered from the cache.
Settings
The Settings page under Logging is where you set up log forwarding. Forwarding sends your application and system logs to a Syslog endpoint you control, so you can keep them longer than 5 days or feed them into your own monitoring tools. See Log Forwarding for the setup steps.
HTTP logs are not forwarded from here. They belong to the pull zone that serves the endpoint, so forwarding for them is configured on the pull zone itself. See CDN log forwarding for details. Last modified on September 2, 2026
