What is the definition of each counter?

Critical Counters

These counters help identify configuration/setup problems that result in data loss, so they're the most critical counters to monitor and correct.

Counter Name Description

movies_discarded_count

desired value: 0

The number of 12-second video segments that have been discarded even before analysis. Any number greater than zero means that the Box gateway is overloaded by having either:
a) too many streams connected to it or
b) by having too-high bitrate/resolution relative to its number of connected streams.

When the Box cannot keep up with the incoming streams, it starts discarding video segments arbitrarily.

events_purged_count

desired value: 0

The number of Events that couldn't be uploaded and indexed before they were purged from disc on the Box gateway. Any number greater than zero means that uplink bandwidth was so severely constrained that not even the Event metadata (without movies) could be uploaded.

movies_purged_count

desired value: 0

The number of 12-second video segments that were missing from the local storage of the Box gateway at the time they were requested for upload to the cloud. Any number greater than zero means that either:
a) uplink bandwidth is insufficient to transfer the video to the cloud, or
b) the on-Box storage buffer is too small to retain the desired amount of video history for lazy retrieval (e.g. the user requested retrieval with the query [7am to 8am 10 days ago], but the Box's storage capacity retains only the 7 days before overwritten by newly arriving video)

segmentation_process_restart_count

desired value: < 2

The number of times that Box had to re-acquire a connection to the video stream after a failure in reading from it. The Box "segments" the incoming video stream into 12-second chunks as the first part of the processing pipeline. Normally, that segmentation process runs continuously without interruption. Any number greater than zero means that:
a) the camera is struggling to serve the video stream reliably, perhaps due to concurrent requests for the same stream by other NVRs or VMS clients, or
b) the local network is unstable and unable to server the video stream reliably on the local network.

The restart is very fast (less than 1 second), so a small number of restarts upon network "glitches" is OK.

Have more questions? Submit a request

Comments