Due to a bug in `nginx-ts-module`, on very long runs (>24h), the HLS/DASH generation seems to stop for a few minutes. Our production workaround is currently to restart HLS/DASH at 3AM, resulting in a few seconds of unavailability.
Due to a bug in
nginx-ts-module, on very long runs (>24h), the HLS/DASH generation seems to stop for a few minutes.Our production workaround is currently to restart HLS/DASH at 3AM, resulting in a few seconds of unavailability.