This chart deploys Teknoir AI NVRs to a Kubernetes cluster.
The implementation of the Helm chart is right now the bare minimum to get it to work.
Use the HelmChart to deploy the Teknoir AI NVR to a Device.
---
apiVersion: helm.cattle.io/v1
kind: HelmChart
metadata:
name: nvr
namespace: default
spec:
repo: https://teknoir.github.io/nvr-helm
chart: nvr
targetNamespace: default
valuesContent: |-
# Example for minimal configuration
instances:
- name: camera-default-name
# Values possible for Teknoir AI NVRs can be overridden for each instance
defaults:
name: camera-default-name
camera:
type: camera
uri: rtsp://host:port/path
# GST pipeline variable substitution:
# - uri: nvr.camera.uri
# - camera_id: release full name | fullnameOverride
pipeline: videotestsrc ! queue ! nvvideoconvert ! video/x-raw(memory:NVMM),width=1920,height=1080 ! tee name=t t. ! nvvideoconvert ! video/x-raw(memory:NVMM),width=[1,1920],height=[1,1080],pixel-aspect-ratio=1/1 ! nvvideoconvert ! queue ! nvjpegenc quality=55 ! appsink sync=0 name=mqttsink t. ! nvvideoconvert ! video/x-raw,format=RGB,width=[1,640],height=[1,360],pixel-aspect-ratio=1/1 ! nvvideoconvert ! motioncells threshold=0.0001 gridx=32 gridy=18 ! fakesink
# The path to the models directory
modelRepositoryPath: /opt/teknoir/models
# The path to the video storage
videoStoragePath: /opt/teknoir/video/segments
# This sets the container image more information can be found here: https://kubernetes.io/docs/concepts/containers/images/
image:
repository: us-docker.pkg.dev/teknoir/gcr.io/nvr
# Overrides the image tag whose default is "ds7-gst1.24.7".
# ds7-gst1.24.7: Nvidia Ada Lovelace GPU on AMD64
# ds7-gst1.24.7-l4t: Nvidia Orin(tegra) on ARM64
tag: ds7-gst1.24.7
# Set the log level for the nvr
# - DEBUG
# - INFO
# - WARNING
# - ERROR
debugLevel: INFO
mqtt:
# Local MQTT broker
broker: mqtt.kube-system
port: 1883
qos: 2
keepAliveInterval: 10
connectRetryInterval: 10
# Topic prefix for generated topic <topicPrefix>/<camera_id: release full name | fullnameOverride>
topicPrefix: camera/frames
# Number of frames to queue before dropping
queueLimit: 15
# Send metrics collected through MQTT
metrics: true
metricsTopic: toe/state
# Send events without inference calculated
nonInferenceEvents: true
# Send events without motion detected
nonMotionEvents: true
motion:
# Number of frames to skip inference when no motion detected
inactiveInterval: 900
# Number of frames to skip inference when motion is detected
activeInterval: 0
healthz:
fpsThreshold: 0.01
resources:
limits:
cpu: 2000m
memory: 2048Mi
requests:
cpu: 100m
memory: 128Mi
nvdsanalytics:
enabled: false
# Mount path for the nvdsanalytics config file: <mountPath>/config_nvdsanalytics.txt
mountPath: /app/config
config: |
[property]
enable=1
config-width=1920
config-height=1080
osd-mode=2
display-font-size=25
[line-crossing-stream-0]
enable=1
line-crossing-Entry-0=59;591;153;557;200;688;106;426
line-crossing-Exit-0=247;523;153;557;106;426;200;688
line-crossing-Entry-1=1607;566;1508;550;1568;183;1448;917
line-crossing-Exit-1=1410;534;1508;550;1448;917;1568;183
# Deprecated in favor of ORG TREE
location:
country: "us"
region: "texas"
branch: "houston"
facility: "office"
site: "interior"
zone: "test"
group: "poc"---
apiVersion: helm.cattle.io/v1
kind: HelmChart
metadata:
name: nvr
namespace: default
spec:
repo: https://teknoir.github.io/nvr-helm
chart: nvr
targetNamespace: default
valuesContent: |-
defaults:
camera:
pipeline: rtspsrc location={{.uri}} latency=1000 ! queue ! rtpjitterbuffer latency=250 ! queue ! rtph264depay ! queue ! nvv4l2decoder ! queue ! nvvideoconvert ! video/x-raw(memory:NVMM),width=1920,height=1080 ! m.sink_0 nvstreammux name=m batch-size=4 width=1920 height=1080 ! queue ! nvinferserver name=nvis config-file-path=/models/rtdetr_config.pbtxt unique-id=2 interval=15 ! queue ! nvinferserver config-file-path=/models/up_down_classifier_config.pbtxt unique-id=5 ! queue ! nvtracker tracker-width=640 tracker-height=384 gpu-id=0 ll-lib-file=/opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so ll-config-file=/models/config_tracker_NvDeepSORT.yml compute-hw=2 ! nvstreamdemux name=d d.src_0 ! queue ! tee name=t t. ! queue ! nvvideoconvert ! video/x-raw(memory:NVMM),width=1920,height=1080 ! nvv4l2h265enc copy-meta=true iframeinterval=15 bitrate=250000 ! queue ! h265parse ! queue ! splitmuxsink sink=nvdsfilesink muxer-properties=properties,streamable=true location=/app/videos/{{.camera_id}}_%05d.mp4 max-size-time=15000000000 max-files=5760 t. ! queue ! nvdsanalytics config-file=/app/config/config_nvdsanalytics.txt ! queue ! nvdsosd ! nvvideoconvert ! video/x-raw(memory:NVMM),width=[1,1920],height=[1,1080],pixel-aspect-ratio=1/1 ! nvvideoconvert ! queue ! nvjpegenc quality=45 ! appsink sync=0 name=mqttsink t. ! nvvideoconvert ! video/x-raw,format=RGB,width=[1,640],height=[1,360],pixel-aspect-ratio=1/1 ! nvvideoconvert ! motioncells threshold=0.00005 gridx=32 gridy=18 ! fakesink
instances:
- name: teknoir-nvr-2x3090-se-axis
camera:
uri: rtsp://teknoir:[email protected]/axis-media/media.amp?videocodec=h264&fps=15&resolution=1920x1080
nvdsanalytics:
enabled: true
config: |
[property]
enable=1
config-width=1920
config-height=1080
osd-mode=2
display-font-size=25
[line-crossing-stream-0]
enable=1
line-crossing-Entry-0=59;591;153;557;200;688;106;426
line-crossing-Exit-0=247;523;153;557;106;426;200;688
line-crossing-Entry-1=1607;566;1508;550;1568;183;1448;917
line-crossing-Exit-1=1410;534;1508;550;1448;917;1568;183
- name: teknoir-nvr-2x3090-se-anders-test
camera:
uri: rtsp://rtsp-video-file-streams:8554/vid5
nvdsanalytics:
enabled: true
config: |
[property]
enable=1
config-width=1920
config-height=1080
osd-mode=0
display-font-size=25helm repo add teknoir-nvr https://teknoir.github.io/nvr-helm/helm install nvr teknoir-nvr/nvr -f values.yaml
