Help... Zoneminder is very slow
Posted: Thu Dec 16, 2021 1:39 pm
We are having frequent dropouts and artifacts on live feed and recorded clips. This happens mostly when there is a lot of movement in front of the cameras.
When we log in to the webinterface of the individual cameras (Hikvision), the feeds are normal and in good quality. The live feeds and recordings in Zoneminder are choppy and in very low FPS (1 to 3 fps). Idle it runs at around 4 to 5 fps. The live feeds also are always 15 to 30 seconds behind the real live events.
We already changed the source method from TCP to UDP, resulting in even worse streams. Resolution of the cameras is set to 1280 x 1024. We had them running at 320 x 240 for a while, but this did not improve anything. Camera function is set to Modect but changing to Monitor does not help.
Cameras and server have their own separate VLAN. Cabling is copper CAT6a.
Video writer is set to X264 Encode (H264 Camera Passthrough was worse). Buffers are set as follows:
Image buffer size: 20
Warmup frames: 5
Pre event image count: 5
Post event image count: 5
Stream replay image buffer: 0
Alarm frame count: 2
The setup is a Ubuntu server 18.04.5 LTS headless running with 6 GB of memory in Hyper-V (Windows Server 2019). There a 6 cameras connected. Storage space is 60 GB (46 GB used) mounted on /mnt/zmdata/
In Hyper-V manager CPU usage is at 9%, HTOP shows values near 100% (user www-data in top place). Memory usage in HTOP is only 2.12GB usage and swp is at 268K / 1.98GB.
The HTOP CPU values are concerning. Strange that Hyper-V shows an almost idle machine. Is there something wrong in my config or does anyone have some other tips that could help me with these problems?
Thanks,
Jef
When we log in to the webinterface of the individual cameras (Hikvision), the feeds are normal and in good quality. The live feeds and recordings in Zoneminder are choppy and in very low FPS (1 to 3 fps). Idle it runs at around 4 to 5 fps. The live feeds also are always 15 to 30 seconds behind the real live events.
We already changed the source method from TCP to UDP, resulting in even worse streams. Resolution of the cameras is set to 1280 x 1024. We had them running at 320 x 240 for a while, but this did not improve anything. Camera function is set to Modect but changing to Monitor does not help.
Cameras and server have their own separate VLAN. Cabling is copper CAT6a.
Video writer is set to X264 Encode (H264 Camera Passthrough was worse). Buffers are set as follows:
Image buffer size: 20
Warmup frames: 5
Pre event image count: 5
Post event image count: 5
Stream replay image buffer: 0
Alarm frame count: 2
The setup is a Ubuntu server 18.04.5 LTS headless running with 6 GB of memory in Hyper-V (Windows Server 2019). There a 6 cameras connected. Storage space is 60 GB (46 GB used) mounted on /mnt/zmdata/
In Hyper-V manager CPU usage is at 9%, HTOP shows values near 100% (user www-data in top place). Memory usage in HTOP is only 2.12GB usage and swp is at 268K / 1.98GB.
The HTOP CPU values are concerning. Strange that Hyper-V shows an almost idle machine. Is there something wrong in my config or does anyone have some other tips that could help me with these problems?
Thanks,
Jef