Monitoring Workflow Executions
Keep track of your workflow runs and catch issues early.
Overview
An execution is a single run of a workflow, triggered manually, by schedule, by webhook, or by email. Every execution is logged with its status, duration, trigger source, and per-node output. The Executions page is your central view of what has run, what succeeded, and what failed across the workspace.
Monitoring matters because workflows fail for reasons that are not always obvious: an OAuth token expires, a third-party API rate-limits, a webhook payload changes shape. Keeping an eye on execution health lets you catch these issues before they cause downstream problems.
Execution Overview
Open Executions from the sidebar to see all workflow runs across your workspace. Each row shows the workflow name, status (Running, Succeeded, Failed, Cancelled), trigger type, start time, and duration. Click any row to drill into the per-node detail view, where you can inspect inputs, outputs, and errors for each node that ran.
Real-Time Monitoring
In the workflow designer, click Run to watch an execution live. Nodes change colour as they progress: blue while running, green on success, red on error. Click a node mid-run to see the input it received; click after completion to see the output. This is the fastest way to debug a workflow as you build it.
Filtering Executions
- Workflow - narrow to a single workflow's history.
- Status - find only failures, or only successes.
- Trigger type - separate scheduled runs from webhook-driven ones.
- Date range - focus on a specific incident window.
Setting Up Alerts
For active monitoring, add a Send Email node, a slack Connector node, or a connector for your incident tool inside a Condition branch that handles the error path. Wrap risky steps in retry settings so transient failures recover without alerting; reserve alerts for the cases that genuinely need a human.
Checking Health
Sweep through execution logs regularly to catch issues that do not surface as outright failures:
- Climbing error rates on a previously stable workflow.
- Execution durations growing over time (often a sign of pagination or list-size creep).
- Repeated authentication failures pointing to an expiring OAuth token.
- Webhook executions dropping to zero, suggesting an upstream system has stopped calling.
Tips
- Pin the Executions tab in your browser during a deploy or schema migration so you can spot failures as they happen.
- Use the workflow filter to compare today's run pattern against last week's.
- Treat a sudden drop in execution volume as suspicious - missing runs are easier to overlook than failed ones.
Common Pitfalls
- Manual runs from the designer also appear in the Executions list. Filter by trigger type if you only want production runs.
- Execution logs age out after the standard retention window. Export anything you need for long-term records.
- A "Succeeded" status only means no node errored. A workflow can succeed while doing the wrong thing - validate outputs as well as status.
Related Articles
- Understanding Execution Logs
- Understanding Retry and Error Handling
- Troubleshooting Connection Issues
- Enabling and Disabling Workflows