system-logs

Compact guide for reading backend + shadow-cljs logs from the monitoring system.

Use When

• •User asks for build/compile status, runtime errors, auth/db issues, or “what just happened”.

Prereqs

• •Monitoring script exists: ./scripts/sh/monitoring/read_output.sh
• •Logs live at /tmp/command_output_{RUN_ID}.log (pointers .pos/.last).

Fast Path

1. •Check status: [ -f /tmp/active_command.txt ] && echo active || echo inactive
2. •Read once: ./scripts/sh/monitoring/read_output.sh
3. •Live follow: ./scripts/sh/monitoring/read_output.sh -f (Ctrl+C to stop)
4. •Need fresh BE/FE compile logs? Restart system over nREPL (app already running):
   • •clj-nrepl-eval --discover-ports → pick the clj port (often 7888).
   • •clj-nrepl-eval -p <port> "(require 'system.core :reload) (system.core/restart-system)"
   • •Add --with-sudo if ps is blocked.

Minimal Workflow

• •Target: server | compile | auth | db | perf | recent | errors.
• •Read: ./scripts/sh/monitoring/read_output.sh (or -f).
• •Filter: pipe to grep -i "error|exception|failed|fatal", "compil|build|warning", "auth|login|token|unauth", or tail -50 for recency.
• •Summarize: list findings and next actions.

If Monitoring Missing

• •Ask teammate to start monitoring run (e.g., via monitor_terminal.sh '<cmd>') then rerun read_output.

Output Template

## Log Analysis
- Source/time: ...
- Errors/warnings: ...
- Perf/DB: ...
- Recent activity: ...
- Next actions: 1) ... 2) ...

Tips

• •Use grep -C 3 for context.
• •Hot-reload/build issues are near the end of the file.