I've got a simple BI setup running on a local server. Not that it should matter, but it's an HP i5-6500 with 40GB RAM, recording to a single 10TB WD Enterprise drive (shucked EMAZ drive). Win10 Pro. BI Version 22.214.171.124 x64 Cameras are a pair of Dahua cameras, models don't particularly matter. They're powered by a Dell 5548P switch, which is on a 1000VA UPS. Location is a commercial building with good power. Network traffic is handled by a Ubiquiti EdgeRouter4. Cameras are on a VLAN that contains only the cameras and one port of the server's 4-port Intel NIC. There is no internet access on this VLAN. The server does have internet access via another port on the NIC. At random, both cameras will drop out with "No Signal" messages on screen and not come back to BI without either restarting BI or rebooting (effectively the same thing). I can login to the cameras at any time, no power cycling required. I have noticed that when this problem occurs, BI is using a ton of CPU (>50%), and when it's restarted, it comes back using ~20%, as expected. BI video settings: 20FPS, direct to disk recording via Intel. Detect loss of singal is checked, as is interpret monochrome video as loss of signal. Nothing else in the Watchdog window is checked. Receive buffer is 40MB on both cameras. I can post screencaps or logs, as required. The main log recorded as: 0 9/8/2019 9:36:31 PM New Move: nothing to do [77/168 hrs, 8.82G/10.0G, 9.05T free] 0 9/8/2019 9:36:31 PM Stored Delete: nothing to do [149/336 hrs, 18.7G/20.0G, 9.05T free] 1 9/8/2019 9:40:36 PM Eyeball Signal: network retry 1 9/8/2019 9:40:36 PM Front Signal: network retry 1 9/8/2019 9:41:06 PM Eyeball Signal: network retry 1 9/8/2019 9:41:06 PM Front Signal: network retry Signal failure continues after that. It's odd that both cameras drop out at the same time. The server is set to never sleep, and never turn off the drives. Any suggestions about how to avoid this in the future? I'm in the process of installing another 4 cameras, and never had this issue prior to BI v5.