Skip to content

Services and the Log Panel

The log panel is where you watch services run. The goal is to get the right logs in front of you — and in front of the agent — fast.

Drag a service from the sidebar into the workspace. An empty panel reads “Drag a service to view logs.” Once a service is in, the panel streams its output.

  • Live means you are following new output as it arrives. Scroll up and following pauses; a ”↓ N new logs” chip lets you jump back to the bottom.
  • Scrolling to the top loads history, with a “history · live below” separator marking where past output ends and the live stream begins.

One bug usually spans more than one service. Open several at once and compare side by side.

From a panel’s toolbar:

  • Split right / Split down to add a panel next to or below the current one.
  • Close panel to remove it.
  • The workspace header offers Balanced vs Columns layout, and maximize / restore the whole workspace.

There is a maximum number of panels; if you hit it, close one before adding another.

Each panel header surfaces counts so you don’t have to read every line:

  • ● N errors and ● N warnings.
  • Filtered N (hidden by your filters) and Folded N (collapsed repeats).
  • For remote services, a node summary: nodes ready/total · collectors, with per-node health (tunnel disconnected, collector missing / stopped / failed).

Type in the panel’s filter box and press Enter to add a keyword. Each keyword is Include or Exclude. You can also select text in a log line and add it to the filter directly.

Inline filters are throwaway. When a filter is worth keeping, save it as a rule — see Log investigation techniques.

The bottom bar shows and controls how wide your view is — which services are in the panels and how many nodes are selected — plus restart / stop on the selected deployment and the evidence pins you have collected (copy / export).