Feeds Go Dead After Awhile

MIP
, ,
1

I can’t be sure if this is an issue with XProtect’s recording server, AI Bridge, or something between them, but I’ve noticed that sometimes, after running for, say, 2.5-3 hours, AI Bridge feeds go dead. Sometimes it’s caused by the recording server going dead, but sometimes the recording server is still going but I can’t access the feeds. A restart of AI Bridge and/or the recording server gets things back on track.

I’m running AI Bridge in a K8s cluster along with an IVA. I’ve gone through the pod logs and nothing jumps out at me as being an error. I’m pulling some 16 feeds concurrently for the duration.

I’m running this setup on a single appliance with plenty of RAM, a decent CPU, and an NVIDIA A2 GPU. Everything seems to swim along just fine until the above happens.

Any ideas as to where I can look to troubleshoot this further?

Thanks in advance.

2

If the Recording Server sometimes goes dead, then I would definitely first try to solve that issue. Even an “half-dead” Recording Server can cause unpredictable results to clients connecting to it.

Maybe you can check the log files here:

[C:\ProgramData\Milestone\XProtect](file:C:/ProgramData/Milestone/XProtect) Recording Server\Logs

And see if there is anything suspicious. I would start looking at the “RecorderEngine.log” and also the “DeviceHandling.log”.

3

Thanks, John. I’ll be sure to check that out.