Get phone and SMS alerts fromPrefect
Prefect orchestrates your data pipelines and scheduled jobs, and tells you when a flow run fails, crashes or runs late. Spike turns that notification into a page for whoever is on call, escalating until someone picks it up.
Set up with AI
I use Prefect for monitoring and Spike for on-call and incident management. Walk me through connecting them: create a Prefect integration in Spike, copy its webhook URL, and add that webhook to Prefect alert settings. Then help me send a test alert to confirm incidents show up in Spike.
What is Prefect?
Prefect is a workflow orchestration platform for data and automation teams. You write flows in Python, deploy them to run on a schedule or on demand, and Prefect tracks every flow run through its states: scheduled, running, failed, crashed, late, completed and so on. When a pipeline breaks overnight, Prefect knows, but a notification in a channel nobody is watching won’t wake anyone.
Pointed at Spike through a Prefect automation and webhook, a firing automation becomes a page: the on-call engineer is called and texted, and an alert nobody acknowledges climbs the escalation policy until someone answers. The incident title is Prefect’s message about the run’s state plus the flow and flow run it happened on, or the automation’s name when Prefect sends no message, for example “Nightly ETL still running after 30 minutes on warehouse-etl/crimson-fox”, so the cause is readable on a lock screen.
An incident is one automation on one flow run. Another delivery from the same automation for the same flow run joins the open incident instead of paging again, and a different flow run or a different automation is a separate incident. The state, the event, the link to the flow run in the Prefect UI and the time it happened ride along on the incident, and alert routing rules can read any field in the payload.
Spike does not auto-resolve Prefect incidents yet. Prefect’s automation form offers no action that runs when an automation resolves, so incidents are resolved in Spike by hand: the responder resolves the incident once the flow run is healthy again.
How to set up Prefect alerts
Every Spike integration is a unique webhook URL. Paste it into Prefect and alerts start flowing — no agents, no code.
Step 1: Create a Prefect integration in Spike
In your Spike dashboard, open Integrations and choose Prefect. Spike generates a unique webhook URL for this integration — copy it, and attach the integration to the service and escalation policy that should answer these alerts.
Step 2: Add a webhook block in Prefect
In Prefect, create a Webhook block with method POST that points at the Spike URL, and set the header {"Content-Type": "application/json"}. Prefect sends no content type of its own, and Spike cannot read the body without it.
Step 3: Add the webhook to an automation
Create an automation whose trigger catches the problem, such as a flow run that stays in a state past a time limit. Under its actions, add Call a webhook with Spike's JSON template and "status": "Triggered".
Step 4: Watch a failed run become an incident
When the automation fires for a flow run, Spike opens an incident titled with Prefect's state message and the flow and flow run it came from. Spike does not auto-resolve Prefect incidents yet, so the responder resolves the incident in Spike once the run is healthy.
That's it — Prefect alerts now page whoever is on call.
