Bandwidth problems recognisable from buffer utilisation

The buffer load of the bridge is a good indicator to detect problems at an early stage.
Normally, the bridge stores the video data on the internal hard disk of the bridge and transfers it to the cloud in a timely manner. Only in cases of malfunction does the buffer load/memory consumption of the internal hard disk increase. This can be the case, for example, if:

  • The internet connection is temporarily down
  • The upload bandwidth is temporarily more heavily utilised.

Negative examples of higher buffer usage are:

  • The internet bandwidth (upload) is not sufficient for the camera data. Often this is due to the cameras producing too many false alarms and consuming significantly more bandwidth than originally planned.

  • Firewall restrictions that block some ports necessary for full communication with the cloud, preventing local data from being transferred to the cloud.

  • Poor internet connection that does not provide stable minimum bandwidth.

  • Non-optimal configuration of the bridge selected due to insufficient bandwidth for data transmission.

  • Transmission mode in the camera cloud settings in the tab “Resolution” only set to demand instead of permanently in the background.

In the Bridge settings of the Cloud web interface, this utilisation can be displayed graphically over time under the tab “Metrics” in the “Storage” area.

Normally, the utilisation of the buffer should show a horizontal “line” that should be below 50 GByte.

The screenshot below shows an unusual buffer load, which clearly indicates that too much data is being buffered. A clearer sign that too much data is being produced by the cameras and cannot be synchronised with the cloud. In this case, reduce the data, check the system for the number of false alarms per camera (e.g. Installer Tool) and consider that a change does not lead to an immediate reduction of the buffer and that this takes some time (depending on the available upload bandwidth).

Especially the transmission mode can be shown well in the follow metrics (CamCloud-BW). This leads to an enormous buffer increase because the cameras permanently synchronise their data with the cloud: