Currently there is a bug where the Transform will say it is unhealthy when it is running successfully.

If the transform encounters an error searching the source index, it will record an error and correctly report itself as unhealthy. Once the transform retries and is able to successfully search the source index, if there have been no documents added to the source index since the last successful checkpoint, then the transform will exit early but it will not reset its health status. The transform is successfully running, but the health data becomes stale.

A workaround is to manually reset the health status by stopping the transform and starting it again:

1. POST /_transform/{transform_id}/_stop
2. POST /_transform/{transform_id}/_start