Failed to create camera provider from aws greengrass lambda with CSI

Hello,

Invoked ‘deepstream-app’ in a greengrass lambda but failed to create camera provider. Any hints will be appreciated.

Below is the command line invoked by python:

bash_command = "./deepstream-app -c ./source1_csi_dec_infer_resnet_fp16_localpath.txt"
process = subprocess.Popen(bash_command.split(), stdout=subprocess.PIPE)
output, error = process.communicate()
print(output, error)

Below is the log file:

[2020-08-02T11:41:05.03+09:00][DEBUG]-IoTDataPlane.py:126,Publishing message on topic “greengrass-deepstream-app/heartbeat” with Payload “RTSP Heart beat signal sent from Greengrass Core running on pla
tform: Linux-4.9.140-tegra-aarch64-with-Ubuntu-18.04-bionic”
[2020-08-02T11:41:05.031+09:00][DEBUG]-Lambda.py:96,Invoking Lambda function “arn:aws:lambda:::function:GGRouter” with Greengrass Message “RTSP Heart beat signal sent from Greengrass Core running on pl
atform: Linux-4.9.140-tegra-aarch64-with-Ubuntu-18.04-bionic”
[2020-08-02T11:41:05.031+09:00][INFO]-ipc_client.py:167,Posting work for function [arn:aws:lambda:::function:GGRouter] to http://localhost:8000/2016-11-01/functions/arn:aws:lambda:::function:GGRouter
[2020-08-02T11:41:05.036+09:00][INFO]-ipc_client.py:177,Work posted with invocation id [d169502c-0bb3-4711-4a5a-f9ce021b3bc8]
[2020-08-02T11:41:06.209+09:00][ERROR]-2020-08-02 11:41:06.209408: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.2
[2020-08-02T11:41:06.976+09:00][ERROR]-(deepstream-app:10): GLib-GObject-WARNING : 11:41:06.976: g_object_set_is_valid_property: object class ‘GstNvArgusCameraSrc’ has no property named ‘maxperf’
[2020-08-02T11:41:11.415+09:00][ERROR]-0:00:06.326112099 10 0x4d0f2f0 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[
UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1577> [UID = 1]: deserialized trt engine from :/lambda/Primary_Detector/resnet10.caffemodel_b1_gpu0_fp1
6.engine
[2020-08-02T11:41:11.415+09:00][ERROR]-0:00:06.326323926 10 0x4d0f2f0 INFO nvinfer gstnvinfer.cpp:602:gst_nvinfer_logger:<primary_gie> NvDsInferContext[
UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1681> [UID = 1]: Use deserialized engine model: /lambda/Primary_Detector/resnet10.caffemodel_b1_gpu0_fp16.en
gine
[2020-08-02T11:41:11.588+09:00][ERROR]-0:00:06.499514536 10 0x4d0f2f0 INFO nvinfer gstnvinfer_impl.cpp:311:notifyLoadModelStatus:<primary_gie> [UID 1]:
Load new model:/lambda/config_infer_localpath.txt sucessfully
[2020-08-02T11:41:11.591+09:00][ERROR]-(Argus) Error FileOperationFailed: Connecting to nvargus-daemon failed: No such file or directory (in src/rpc/socket/client/SocketClientDispatch.cpp, function ope
nSocketConnection(), line 201)
[2020-08-02T11:41:11.591+09:00][ERROR]-(Argus) Error FileOperationFailed: Cannot create camera provider (in src/rpc/socket/client/SocketClientDispatch.cpp, function createCameraProvider(), line 102)
[2020-08-02T11:41:11.591+09:00][ERROR]-Error generated. /dvs/git/dirty/git-master_linux/multimedia/nvgstreamer/gst-nvarguscamera/gstnvarguscamerasrc.cpp, execute:551 Failed to create CameraProvider
[2020-08-02T11:41:11.599+09:00][ERROR]-NvMMLiteOpen : Block : BlockType = 4
[2020-08-02T11:41:11.6+09:00][ERROR]-===== NVMEDIA: NVENC =====
[2020-08-02T11:41:11.601+09:00][ERROR]-NvMMLiteBlockCreate : Block : BlockType = 4
[2020-08-02T11:41:13.01+09:00][INFO]-run_deepstream.py:46,('\n *** DeepStream: Launched RTSP Streaming at rtsp://localhost:8554/ds-test \n\nOpening in BLOCKING MODE \nINFO: [Implicit Engine Info]: l
ayers num: 3\n0 INPUT kFLOAT input_1 3x368x640 \n1 OUTPUT kFLOAT conv2d_bbox 16x23x40 \n2 OUTPUT kFLOAT conv2d_cov/Sigmoid 4x23x40 \n\n\nRuntime commands:\n\th:
Print this help\n\tq: Quit\n\n\tp: Pause\n\tr: Resume\n\nNOTE: To expand a source in the 2D tiled display and view object details, left-click on the source.\n To go back to the tiled display, righ
t-click anywhere on the window.\n\n
INFO: <bus_callback:181>: Pipeline ready\n\n
INFO: <bus_callback:167>: Pipeline running\n\n
* INFO: <bus_callback:204>: Received EOS. Exiting …\n\nQuitting\nAp
p run successful\n’, None)
[2020-08-02T11:41:13.011+09:00][INFO]-lambda_runtime.py:147,Running [arn:aws:lambda:ap-northeast-1:042083552617:function:greengrass-deepstream-04208355-DeepstreamAppLambda-1QIZMKLLSX9P5:9]
[2020-08-02T11:41:13.011+09:00][INFO]-ipc_client.py:192,Getting work for function [arn:aws:lambda:ap-northeast-1:042083552617:function:greengrass-deepstream-04208355-DeepstreamAppLambda-1QIZMKLLSX9P5:9
] from http://localhost:8000/2016-11-01/functions/arn:aws:lambda:ap-northeast-1:042083552617:function:greengrass-deepstream-04208355-DeepstreamAppLambda-1QIZMKLLSX9P5:9/work
[2020-08-02T11:41:25.191+09:00][INFO]-lambda_runtime.py:364,Caught signal 15. Stopping runtime.

With the following config files:

config_infer_localpath.txt (3.0 KB)
source1_csi_dec_infer_resnet_fp16_localpath.txt (2.0 KB)

It works well with these 2 configuration files if I execute deepstream-app under ssh or desktop’s terminal.

Hi,
We don’t have experience of using greengrass lambda. Does it work with video file source? Would like to know if it is specific to camera source. And do you use DS4.0.2 or DS5.0 DP on Jetson Nano?

I was using DS5.0 DP on Nano. It works well with video file and rtsp. I don’t have a USB so don’t know that.

I suspect it is this socket - /tmp/argus_socket - according to this post.

However the greengrass deployment fails if I mount it.

Hi,
Does it work if you use greengrass lambda to run gst-launch command like

gst-launch-1.0 nvarguscamerasrc ! fakesink
gst-launch-1.0 nvarguscamerasrc ! fakesink

I will run it later.

I just tested the greengrass lambda with non-container mode as root which was working fine.

Thanks

In container mode on greengrass I got this error below.

[2020-08-06T11:39:33.18+09:00][ERROR]-Error generated. /dvs/git/dirty/git-master_linux/multimedia/nvgstreamer/gst-nvarguscamera/gstnvarguscamerasrc.cpp, execute:568 Failed to create CaptureSession

Hi,
Please enable user-space logs and attach it for reference.
https://elinux.org/Jetson_TX2_Camera_BringUp#Steps_to_enable_more_debug_messages