For the complete documentation index, see llms.txt. This page is also available as Markdown.

Logs

Logs is the raw output of your deployed containers: agents, services, and databases. Where Observability records what an agent decided and Analytics records how its container performed, Logs is what the process actually printed.

Pick a workload and a time range and the stream updates on whatever refresh interval you set. History reaches back 72 hours.

Reading a line

Each line carries a timestamp, a level, and a message. The level filter counts what it finds as well as filtering, so you can see there are thirty-seven errors in the window before deciding whether to open any of them.

Opening a line shows the structured fields behind it. Every line carries a core set: trace id, level, timestamp, message, the logger that emitted it, and the function and line number it came from. Beyond that a line carries whatever the code attached, so a routine evaluation arrives with its routine, node count, and duration as separate fields rather than buried in the message text. The JSON Fields column counts them, and the whole payload is available raw as JSON.

Logs is container output. To see what an agent did inside a request, see Observability. For traffic, latency, and resource use, see Analytics.

Last updated

Was this helpful?