Skip to main content
This article explains why CloudTalk automatically disables a webhook endpoint that keeps failing, and how to get it running again, including recovering the events you missed.
User level: Admin

​
Why an endpoint gets disabled

When a delivery to your webhook endpoint doesn’t get a successful (2xx) response, CloudTalk retries it automatically, up to 50 times over about 11.5 hours. If your endpoint keeps failing continuously for about 5 days, CloudTalk disables it automatically and stops sending events to it. This keeps a broken endpoint from building up an endless backlog of failed deliveries. Account admins receive an email when an endpoint is disabled, and again when it’s re-enabled.

​
How to get it working again

  1. Fix your endpoint so it returns an HTTP 2xx response to incoming webhook requests.
  2. If you fix it within 12 hours of the endpoint being disabled, there’s nothing more to do. During those 12 hours, CloudTalk keeps testing your endpoint. As soon as it responds successfully, CloudTalk re-enables it automatically and replays the events it missed.
  3. If more than 12 hours have passed, turn the endpoint back on yourself:
    1. Go to Settings > Webhooks and open the endpoint.
    2. Switch the endpoint back on.
    3. Use Recover to replay the deliveries that failed while it was disabled.

​
How long missed events are kept

Event payloads are kept for 30 days. Within that time, you can resend single events or recover failed deliveries. After 30 days, they’re deleted and can no longer be replayed.
If you’re testing webhooks with a public testing tool, such as webhook.site, remember that these endpoints can go offline or stop responding. That counts as a failing endpoint, so it can be disabled the same way as a production endpoint.
For the full retry schedule and tips for building a reliable endpoint, see Webhook Delivery, Retries, and Best Practices.
Need help or have a question? Reach out through our Support portal. We’re here for you.