Every task execution is recorded in the Task Log. Useful for verifying schedules are firing on time, debugging failures, monitoring task duration, and tracking output that tasks produce.
What's recorded
Per log entry:
- Task name — which task ran.
- Start timestamp.
- End timestamp (or null if still running).
- Duration — computed end - start.
- Result — Success, Failed, Timed Out, Skipped.
- Output — any text the handler echoed or returned.
- Error message — if Failed, the error reason.
- Stack trace — if Failed and stack-trace capture is on, the full PHP stack.
- Resources consumed — memory usage, query count (when available).
Where to find it
Architect Panel → Automation → Task Log. The list is filterable by task name, status, and date range. Each entry can be clicked into for the full output and stack trace.
Common queries
1. "Did the nightly task run last night?"
Filter by task name + last 24 hours. You'll see the run (if it happened) with its result. If it's missing entirely, the schedule didn't fire — check the cron entry on the server.
2. "Why did this task error?"
Click into the failed entry. Read the error message and stack trace. Common causes:
- External API down or rate-limited.
- Database lock timeout (long-running task blocked by another query).
- Memory exhausted (task processed too much in one pass).
- PHP exception in the handler code.
- Configuration changed since the task was written.
3. "How long does this task usually take?"
Filter by task name and sort by duration. Spot tasks that are getting slower over time (typically a sign that they're scaling badly with growing data volume).
4. "Which tasks failed in the last week?"
Filter by status = Failed and date range. Useful for end-of-week health reviews — any task that's failing repeatedly is a problem to investigate.
Retention
By default, log entries are kept indefinitely. For very active task installs (hundreds of tasks running every minute), this grows the database significantly over time. Set up a meta-task that purges entries older than your retention policy (typically 30-90 days).
Re-running
From the Task Log entry's detail page, there's a Re-run action. Useful for:
- Manually re-triggering a failed task after fixing the underlying problem.
- Running an ad-hoc execution to test changes.
- Catching up after a planned outage where tasks didn't fire on schedule.
Tip: Configure email alerts on failed tasks. A simple meta-task that runs every 15 minutes and emails an admin when any task in the last interval has failed catches issues quickly. Without an alert, failures can go undetected for days because admins rarely browse the Task Log proactively.