I want to know why some health check transactions still appear even though the Health Check inbound filter is enabled.
​

The Health Check filter only drops transactions whose name matches Sentry’s fixed glob list (for example, *healthcheck*, */health, */ping, */ready, */up). Names outside that list (for example, ping_health) are kept.
​

Full pattern list: Transactions Coming from Health Check.
​

To drop custom names, filter in the SDK with beforeSendTransaction / before_send_transaction (or your SDK’s equivalent). See Filtering.
​

This filter does not drop error events from health check endpoints - see Why am I still seeing health check errors?.