I thought I had fixed it when I ordered another netgear GS108PP unmanaged switch and another 90 watt POE injector for the Hikvision speed dome because as soon I i ordered it the problem stopped. well today we got some rain and it started up again.
in this print screen the car is in 3 different locations. but because of the rain I will have to wait to change out the switch and POE so maybe tomorrow or Sunday View attachment time warp.mp4
I thought I had fixed it when I ordered another netgear GS108PP unmanaged switch and another 90 watt POE injector for the Hikvision speed dome because as soon I i ordered it the problem stopped. well today we got some rain and it started up again.
in this print screen the car is in 3 different locations. but because of the rain I will have to wait to change out the switch and POE so maybe tomorrow or Sunday View attachment 190245
I noticed this problem within the past year or so. This issue seemed to only affect cloned cameras. When a camera that has a clone reboots or drops off the network, the second or "cloned" camera in BI will stop responding when the camera comes back online. Rebooting the BI computer will solve the issue, but the problem will come back when a camera disconnects and re-connects.
@Coldair 5 of your cameras in that clip are frozen since almost a minute before you started recording that screen capture. See the timestamps.
I suspect something in Blue Iris's code for creating an alert or clip is hanging up waiting for something that wasn't supposed to take any time at all, and stalling the entire video rendering pipeline for the affected cameras. It happens rarely on my system, rarely enough that I haven't been able to really investigate. I couldn't even tell you if the clips it was trying to record at the moment of the freeze are okay or if they also freeze at that spot. The live view always unfreezes at about the time when the clip would stop recording anyway (so within 15-30 seconds on my system).
I did not pay attention to that, but what is really odd was it had been working pretty well so much so I was going to contact support and have them install Ai to gut down all the alerts caused by shadows and trees moving in the wind.
the other thing is it happens at random times and with random cameras
After a lot of testing I'm convinced that the problem (at least in my situation) has to do with the CURL commands I'm using to send high resolution still alert images to Pushover, along with the limited upload bandwidth I get from my crappy DSL connection (1Mbps up, 12 Mbps down). Once I have 3 or more cameras alert, and I have 6 cameras that cover our access road, so pretty much every time a vehicle comes in or goes out, I see the behavior. I think that the contention for the upload bandwidth is making the 3rd+ CURL command to exceed some sort of timeout.
If I change the CURL command so that it doesn't attach a JPG, the behavior goes away.
Ken has tried unsuccessfully to reproduce the problem.
For now I told him to not spend any more time on my issue. Yesterday the ISP that is finally deploying fiber in our area ran fiber and installed an OTU at my house. In the next few weeks they should finish splicing the main line and I'll be activating 1 Gbps symetrical service. When that happens I think the problem will go away.
I should then be able to confirm whether or not my suspicions are correct, as my Ubiquiti UDM Pro router has the ability to set up a traffic rule that will allow me to throttle the upload bandwidth from my BI server to the Pushover servers' IP addresses.
Well, as I had expected, going from 1Mbps upload speed to 1Gbps upload speed has made the problem go away for me. We activated my fiber service yesterday, and this morning I upgraded BI to the latest version, v 5.9.0.6. I've had zero instances of the multi camera freeze when several cameras alert and all send notfications containing high rez alert images to Pushover.
I haven't had the chance yet to set up a traffic rule to see whether I can make the problem come back by throttling my upload bandwidth to the Pushover API servers back down to 1Mbps, but I ought to be able to try that out tomorrow.
Hi All ,I am a newbie here. So please excuse my stupidity on IP cameras.
I've been running BI for about 5 years and been running into similar issues with my setup I am running 10 cameras and they all freeze or drop signal randomly. Mostly I think were network glitch's.
9 of them are bullet cams ( cheap ones) and 1 Sunba PTZ Illuminati ,which has been great. Untill I noticed that the bullet cams are seeing action that the Sunba was not. After some investigation I found that during Live streaming the Sunba camera was stalling down to 2 or 3 FPS and the time stamp was stopping 5 to 10 seconds sometimes as much as a minute. Now this has been observed ONLY in BI. For that I assumed that I had a power issue with the Sunba being a PTZ and on POE wire. I changed power source and cable ( NO change )
Now what? So by chance I went to my smartphone BI app and found the same thing lagging Frame rate and jumping time stamp. And poor PTZ control which I also noticed on my desktop. Now for the kicker. If I use the Guarding Vision app I have no issues, Time stamp flow is steady The frame rate at or above 20, PTZ control is fluid and smooth.
NO I have not upgraded to BI 6 yet. It just seems that BI doesn't like the Sunba camera. And has a connection issues with it.
I think its in the network because Guarding vision is a little glitchy on my laptop. but not on my smartphone
Side note: Power surge came in on the internet cable and kill my other 5 year old desktop running BI, but that is another story.
Any thoughts or solutions?
Thank to all this forum is great should have been here sooner.
What computer hardware are you running Blue Iris on exactly? perhaps delete the Sunba and do a new Find Inspect and use defaults to get it running.
There are many settings you could have turned on or off in 5 years lol....hard to say.
Lightning does strange things to devices. sometimes its hard to nail down.
maybe the network switch is glitching out?