Cameras losing connectivity

[PsychloneZA]

Good day

Complete newbie here.

I have Blue Iris (Ver: 5.9.5.2 x64) running on a Win 10 Pro x64 PC (Dell i5-7500, 16GB DDR3) connected to a Mikrotik mANTBox 15s Sector.
I have 4 cameras on a pole connected to a Duxbury 8port PoE switch which is connected to a Mikrotik SXTsq Lite5 CPE (A total of 8 poles.)

The issue:
When Blue Iris is open or the service running, the cameras intermittently show as "No Signal". This is happening with all the cameras.
If I continuous ping a camera while Brue Iris and/or its service is running, the response is 300 - 1200ms
If I continuous ping a camera while Brue Iris and/or its service is NOT running, the response is 1 - 2ms

I have checked the signal strengths on the Mikrotik and they all show good.

I'm guessing the issue is with Blue Iris itself as connectivity to the cameras is only affected when it is running. Any advice and direction on how to resolve this issue would be greatly appreciated.

Kind Regards

Andrew

[Pogo]

I'm guessing the issue is with Blue Iris itself

A careful review of your networking may reveal the problem is the LAN and Blue Iris is just the symptom..., or is incorrectly configured for the actual LAN environment itself.

[PsychloneZA]

Okay, so how do I go about checking the network configuration of Blue Iris? As there are no issues with network connectivity or packet loss with anything else connected on the network and login in to a camera has no issues either, only when Blue Iris starts.

[Pogo]

Good signal strength doesn't necessarily mean good signal quality. And there's more than a little to unpack in your network description and method of testing.

All bets are usually off for most folks when it comes to wi-fi and surveillance devices. I happen to not be part of that crowd and have successfully integrated wi-fi into my camera network for years..., starting by hanging an ethernet -> wifi transceiver off the back of a Swann DVR with eight HD720p cameras to hit a Netgear 6400 router with a couple of WyzeCam v3s and Blue Iris hanging off that. Worked great. LOL Seriously. Easy when you know how...., (or get lucky!).

Anyway, all has advanced somewhat since then and still works as intended both wired and wi-fi. One simply needs to be realistic about their expectations and understand the capabilities of the devices and network configuration to achieve reasonable results.

'Nuff o' that.

I still contend that Blue Iris is simply exhibiting a network issue rather than introducing one. If you are using the BI machine as the source of your testing, have you tried a tracert back to it from the wi-fi segment to possibly see where the latency is introduced?

Have you taken the time to review the entire network architecture for anomalies that may be impacted by introducing the BI network component into the mix? Is the BI web server active?

Any funny ports or security management stuff that would affect access to the Blue Iris internal LAN address? Double checked the IP config of the BI machine? No other network related stuff running on the box? Funky subnetting or VLANs? IP conflicts? Two DHCP servers banging heads? Smart Home stuff in the mix somewhere trying to outsmart everything else?

Wi-Fi interference possibilities? 2.4 or 5gHz. AP or bridge hop mode?

Have you confirmed matching camera configurations between their onboard firmware configurations and the corresponding Blue Iris configurations?

What is the CPU/GPU doing when BI is running? QuickSync enabled? Are you using substreams and doing all the usual optimizations to get the most out of the available hardware and resources? The i5-7500 w/16G of RAM should do just fine when dialed in. On the other hand, what cameras are we talking about? Eight 4K cameras with no optimizations can potentially bring that system to its knees.

Just a few things to consider sorting through.

You'll get it. May just take a bit more head scratching and staring at stuff.

[PsychloneZA]

Hi Pogo

Thank you for the update.

I have been working with the networking, but can't seem to find the issue. From the Blue Iris machine, I can ping all the HikVision cameras with 1ms pings. I have two nANTBox 15s 5GHZ sectors in PTP-Bridge AP mode on the main building connected to a Netgear PoE switch. No VLANS. Eight SXTsq Lite5 CPE's talk to the sectors and each CPE is connected to a Duxbury DUX1010G PoE switch with four Hikvision DS-2CD2086G2-I IP cameras.

The Mikrotiks are all set up with basic configurations. Everything is on static IP's

I can view the web interface live view of the camera when I connect directly to it. I get full video quality with minimal impact on the ping with I continuously ping the camera at the same time. But lose quality and ping when I start Blue Iris or even just the service.

With regards to the camera configs, I am using the preset that Blue Iris provides and have configured but the mainstream and substream info.

The CPU/GPU is chilling out at 15% RAM at 1.6GB. Not using Quicksync

Maybe I need to do some more research and some studies on Mikrotik, though I am still not sure about those Duxbury's... As with the head scratching, that is pretty easy as I have no hair...

[Pogo]

As a real ground floor traffic congestion / camera configuration conflict test, try just disabling all of the cameras at the Blue Iris server level (in their respective video configurations) and then check the impact on the network after restarting Blue Iris. In other words, isolate the cameras themselves from the Blue Iris service. My guess is all will be fine on the network otherwise and pings from the BI box will be closer to the 1ms range with Blue Iris running. That will tell us if the service itself is or isn't introducing the latency.

"I am using the preset that Blue Iris provides and have configured but the mainstream and substream info."

You sorta lost me there. Did you use the Find/Inspect function to scan the cameras, or select them directly from the list, or something else to determine the main and sub links?

** But another weird aspect worth noting when using bridged wi-fi setups is that device MAC addresses behind bridges are often invisible to network resources that would otherwise rely on them to correctly identify (and subsequently manage) said devices. Blue Iris provides a workaround to mitigate that in the camera IP configuration pane as 'Skip initial MAC, HTTP, DNS reachability tests". It is typically an unchecked box and BI uses all the methods indicated to scan devices for legitimacy and ID purposes across the network -- usually just when they are coming online, but...

If the first test of disabling the cameras indicates the BI service itself is cool, you could see what effect disabling all that camera scanning may have on things. Depending on any number of possibilities within your network and its hardware, there may be something else generating huge amounts of device identification requests based on any number of protocols that can't satisfy ID requirements aside from (or along with) Blue Iris -- possibly due to hidden MAC addresses. You do have some fairly capable hardware in the mix that could simply be doing what it's supposed to be doing under such circumstances, but that should also be flexible enough to accommodate a wide variety of scenarios simply by changing configuration parameters.

You have a lot going on. Break it down to the most fundamental elements keeping the troubleshooting as basic as possible until it isn't anymore. That's usually where you'll find the problem. LOL