Moving BI to new PC, everything is going pear-shaped...

wittaj

IPCT Contributor
Joined
Apr 28, 2019
Messages
24,436
Reaction score
47,559
Location
USA
LOL - I couldn't remember who had it LOL.

Yeah, the latest updates seem to not like HA. Some folks here turned HA off on every camera and then slowly added them back until they found the buggy camera or the system got unstable.

With the substream option, the use of HA isn't as critical needed as before. With 50 cameras maybe though LOL.
 

erkme73

BIT Beta Team
Joined
Nov 9, 2014
Messages
1,540
Reaction score
1,412
I think deep stack is also a bit more cpu intensive than people may realize. When it's idle it appears to make no difference. But when there is any analysis occurring, it spikes my cpu into the 90s. Doesn't last long but it happens often when there is any analysis - whether it ends in a confirmed or cancelled alert.

It's a powerful tool, but it drives me nuts to hear it spin up the CPU fan so often and so high.

Eta: it certainly looks like it, since deep stack runs python.Screenshot_20210530-094050.png
 
Last edited:

Flintstone61

Known around here
Joined
Feb 4, 2020
Messages
6,587
Reaction score
10,894
Location
Minnesota USA
Clean install with Win Media Creation Tool was what helped my i5-8500 Elitedesk. Of course im on a different chip than than the 6700....But it's been emailing alerts since Wednesday with out interuption. I used to come in on Mondays and find the service running with a funked up screen with small square patch in the center.
 

Flintstone61

Known around here
Joined
Feb 4, 2020
Messages
6,587
Reaction score
10,894
Location
Minnesota USA
LOL - I couldn't remember who had it LOL.

Yeah, the latest updates seem to not like HA. Some folks here turned HA off on every camera and then slowly added them back until they found the buggy camera or the system got unstable.

With the substream option, the use of HA isn't as critical needed as before. With 50 cameras maybe though LOL.
I did something like that.....Cpu was high so i tried toggling global HA across Intel VPP/Intel/ & Intel Beta....then Used "default" in each cams settings for HA. when i toggled the Jide techs hardware acceleration CPU would go up and down....My demonic Camera is the Jidetech mini PTZ in the Lobby. in this new config anyway. I dropped frames to 7.

watch out for the Jpegs in the Record tab. new defaults!
Cpu hit 99% after I ran thru and enabled [ ] Jpegs in the record tab. ( The newer Versions of BI have new options/ Continous Jpegs was default) LOL ...But i fixed that
 

erkme73

BIT Beta Team
Joined
Nov 9, 2014
Messages
1,540
Reaction score
1,412
Is this normal? I'm seeing two instances of python, both of which use up the balance of my CPU.Screenshot_20210530-150621.png
 

Flintstone61

Known around here
Joined
Feb 4, 2020
Messages
6,587
Reaction score
10,894
Location
Minnesota USA
im not running deepstack (yet). Somebody will answer back soon.
i get enough data from my system right now to address the condo issues. Like yesterday "The Garage door came down on my car! ( yeah because you drove in after the door was previously opened by another car 19 seconds earlier.) And werent paying attention).
 

wittaj

IPCT Contributor
Joined
Apr 28, 2019
Messages
24,436
Reaction score
47,559
Location
USA
I wonder if one python is for objects and the other is for faces?
 

sebastiantombs

Known around here
Joined
Dec 28, 2019
Messages
11,511
Reaction score
27,690
Location
New Jersey
@erkme73 From what I've seen an instance of python is needed for each camera that is triggered and set to use DS. I will say your CPU utilization is kind of high unless you're running 30-40 cameras. Have a look at sub streams to cut it way down.
 
Last edited:

erkme73

BIT Beta Team
Joined
Nov 9, 2014
Messages
1,540
Reaction score
1,412
I wonder if one python is for objects and the other is for faces?
I think it's an instance per camera - at one point I had four of them going. I've disabled DS under global settings and on the cameras.
 

erkme73

BIT Beta Team
Joined
Nov 9, 2014
Messages
1,540
Reaction score
1,412
@erkme73 From what I've seen n instance of python is needed for each camera that is triggered and set to use DS. I will say your CPU utilization is kind of high unless you're running 30-40 cameras. Have a look at sub streams to cut it way down.
That's what I am thinking as well. I am using sub-streams on all cameras. With DS turned off, CPU is now hanging around 30-40% consistently.
 
Top