this is going on for days.
BUT:
what wrong… again?
Hi drsash
The green “healthy” badge and the per-message failures answer different questions. The badge tells you the webhook is still enabled; it doesn’t tell you every delivery landed. So “healthy” plus a run of failed to send webhook in Live data isn’t a contradiction — it just doesn’t tell you why.
The why is in the event. In Live data, click one of the as.webhook.fail entries to expand it. That event carries the HTTP status code the endpoint returned, the URL it called, and up to 1 KB ofthe response body. That usually ends the guessing — in the long-running thread on this, TTN
message “Failed to send webhook” ( TTN message "Failed to send webhook" ), it turned out the receiving API was returning the error itself, and nobody could see it until the metadata was expanded.
If I’m reading your screenshot right this is the TTN Mapper integration rather than an endpoint of your own, which changes what’s actionable. The status code splits it three ways:
- 4xx — the far end is rejecting the payload. That’s usually consistent rather than intermittent, and points at the integration config or the message itself.
- 5xx, or a timeout — the far end is struggling. The Things Stack allows a webhook 5 seconds by default, so a busy community service produces exactly this pattern: mostly fine, intermittent
failures, badge still green. Nothing fixable at your end. - x509_unknown_authority or a DNS error — the network path, not the application.
Worth saying when you post the expanded event: does it fail on every uplink, or only some? Consistent failure points at configuration or authorisation; intermittent points at load or timeouts at the other end. The screenshots can’t distinguish those two, and the fix is completely
different.