Rtsp Sink not working inside docker container

sharan.phadke · October 16, 2023, 6:18am

Please provide complete information as applicable to your setup.

• Hardware Platform Nvidia GPU
• DeepStream Version 3.6
• TensorRT Version
**• NVIDIA GPU Driver Version 530.30.02 **
• Issue Type bug
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
I am Currently trying to run a sample example deepstream application and trying to generate rtsp stream as output. I am running nvidia-deepstream inside docker container

Docker Image used : nvcr.io/nvidia/deepstream:6.3-triton-multiarch
Docker run command used : docker run --gpus all --net=host -it --rm -e DISPLAY=$DISPLAY --device /dev/snd -v /tmp/.X11-unix/:/tmp/.X11-unix -v ./sample-project/:/opt/nvidia/deepstream/deepstream/sources/project -w /opt/nvidia/deepstream/deepstream/sources/project nvcr.io/nvidia/deepstream:6.3-triton-multiarch

I am following steps mentioned in this blog NVIDIA Deepstream Quickstart. Run full YOLOv4 on a Jetson Nano at 22… | by Victor Sonck | ML6team

I have configured the source properly, and the application works perfectly for file sink
But as soon as I change the sink to rtsp (sink=4) the application says the rtsp stream has been started (rtsp://localhost:8554/ds-test) successfully but I cannot access the stream using either vlc or ffmpeg

I have already checked network connectivity (docker container running as net=host), port busy issues, and even tried to access the stream from within the container using gstreamer gst-launch-1.0 uridecodebin uri=rtsp://localhost:8554/ds-test
but still doesn’t work

Everywhere it gives me same error 503 Service Unavailable.

Following is my config file
[application]
enable-perf-measurement=1
perf-measurement-interval-sec=1

[source0]
enable=1
#Type - 1=CameraV4L2 2=URI 3=MultiURI 4=RTSP
type=4
uri=rtsp://localhost:8555/sample
num-sources=1
gpu-id=0
cudadec-memtype=0

[streammux]

live-source=0
batch-size=1

batched-push-timeout=40000

width=1280
height=720

[primary-gie]
enable=1
model-engine-file=…/engines/resnet10.caffemodel_b30_gpu0.engine

batch-size=1
interval=0

gie-unique-id=1
config-file=model_basic.txt

[tiled-display]
enable=1
rows=1
columns=1
width=1280
height=720

[osd]
enable=1
border-width=2
text-size=15
text-color=1;1;1;1;
text-bg-color=0.3;0.3;0.3;1
font=Serif

[sink0]
enable=1
type=3

container=1

codec=1

enc-type=0
sync=0
bitrate=2000000

profile=0
output-file=output.mp4
source-id=0

[sink1]
enable=1

type=4

container=2
#1=h264 2=h265 3=mpeg4

codec=1
sync=0
bitrate=4000000

source-id=0
rtsp-port=8554
udp-port=5400

[tests]
file-loop=0

I have also tried other sample applications provided inside the container and in the documentation, and all of them give the same error, so its probably not an issue with config file.
Is it an issue with the deepstream docker container?

Please help me I am stuck at a deadend

fanzh · October 16, 2023, 8:10am

please enable [sink2] in /opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt, then try deepstream-app -c source4_1080p_dec_infer-resnet_tracker_sgie_tiled_display_int8.txt, then share the output log.
is 8554 taken by other apps?

sharan.phadke · October 16, 2023, 9:17am

The pipeline runs perfectly, I even get the displayed output of Passengers and Cars recognition.
But when I try to play the output rtsp stream (rtsp://localhost:8554/ds-test) using ffplay, It gets stuck for a while and then I get this error

[rtsp @ 0x7f9278000b80] method DESCRIBE failed: 503 Service Unavailable
rtsp://localhost:8554/ds-test: Server returned 5XX Server Error reply

Also I have made sure no other application or process is using the tcp port 8554.

fanzh · October 16, 2023, 10:10am

please share the test log. wondering if there is some special printings.
can you check if the 8554 is take by deepstream-app in container? if yes, can you use gst-discoverer-1.0 or ffprobe to test the rtsp url?
if still can’t work, please try other RTSP port.

sharan.phadke · October 17, 2023, 3:54am

I get these warnings as I start the deepstream-app :

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.094: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_udp.so’: librivermax.so.0: cannot open shared object file: No such file or directory

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.098: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/libgstmpeg2dec.so’: libmpeg2.so.0: cannot open shared object file: No such file or directory

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.100: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/libgstmpg123.so’: libmpg123.so.0: cannot open shared object file: No such file or directory

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.103: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/libgstchromaprint.so’: libavcodec.so.58: cannot open shared object file: No such file or directory

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.127: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/libgstopenmpt.so’: libmpg123.so.0: cannot open shared object file: No such file or directory

(gst-plugin-scanner:114): GStreamer-WARNING **: 03:43:56.152: Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/libgstmpeg2enc.so’: libmpeg2encpp-2.1.so.0: cannot open shared object file: No such file or directory

*** DeepStream: Launched RTSP Streaming at rtsp://localhost:8554/ds-test ***

WARNING: [TRT]: CUDA lazy loading is not enabled. Enabling it can significantly reduce device memory usage. See CUDA_MODULE_LOADING in 1. Introduction — CUDA C Programming Guide
WARNING: …/nvdsinfer/nvdsinfer_model_builder.cpp:1487 Deserialize engine failed because file path: /opt/nvidia/deepstream/deepstream-6.3/samples/configs/deepstream-app/…/…/models/Secondary_CarMake/resnet18.caffemodel_b16_gpu0_int8.engine open error
0:00:02.706939911 113 0x56367f122550 WARN nvinfer gstnvinfer.cpp:679:gst_nvinfer_logger:<secondary_gie_2> NvDsInferContext[UID 6]: Warning from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1976> [UID = 6]: deserialize engine from file :/opt/nvidia/deepstream/deepstream-6.3/samples/configs/deepstream-app/…/…/models/Secondary_CarMake/resnet18.caffemodel_b16_gpu0_int8.engine failed
0:00:02.732119772 113 0x56367f122550 WARN

But it runs perfectly thereafter, displaying PERF of the model, and I even get the popup with car and passengers recognized

The problem remains same when I try to read rtsp stream I get the same error
[rtsp @ 0x7f9278000b80] method DESCRIBE failed: 503 Service Unavailable
rtsp://localhost:8554/ds-test: Server returned 5XX Server Error reply

As I am running the container on host network mode I am pretty sure the port is not used by any other process

sharan.phadke · October 17, 2023, 5:49am

Update: I have also tried running it on different port (8556) but I get the same issue as before

fanzh · October 17, 2023, 6:15am

no narrow down the issue, please do some tests.

when deepstream-app is running, can you use this command to test in container? will the printing refresh?
sudo tcpdump -v host 224.224.255.255
if using filesink, can the output file play? wondering if the encoding is fine.

sharan.phadke · October 17, 2023, 8:56am

The tcpdump command gives this as output
host.docker.internal.51017 > 224.224.255.255.5400: UDP, length 1354
08:52:56.086316 IP (tos 0x0, ttl 1, id 10202, offset 0, flags [DF], proto UDP (17), length 1373)
host.docker.internal.51017 > 224.224.255.255.5400: UDP, length 1345
08:52:56.118051 IP (tos 0x0, ttl 1, id 10204, offset 0, flags [DF], proto UDP (17), length 1428)
host.docker.internal.51017 > 224.224.255.255.5400: UDP, length 1400
08:52:56.118089 IP (tos 0x0, ttl 1, id 10205, offset 0, flags [DF], proto UDP (17), length 273)
and keeps on printing till the app stops

Yes I am unable to play the out.mp4 file generated, when I run app with file sink enabled

fanzh · October 17, 2023, 9:22am

seems it is related to encoding.

what is the GPU device model?
please test this command, then share the 1.log and test.264.
gst-launch-1.0 filesrc location=/opt/nvidia/deepstream/deepstream/samples/streams/sample_720p.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! nvvideoconvert ! video/x-raw$memory:NVMM$,format=I420 ! nvv4l2h264enc bitrate=1000000 ! filesink location=test.264 >1.log 2>1.log

sharan.phadke · October 17, 2023, 9:28am

Nvidia GPU Model : NVIDIA GeForce GTX 1650/PCIe/SSE2

1.log output :
Setting pipeline to PAUSED …
Pipeline is PREROLLING …
Redistribute latency…
Pipeline is PREROLLED …
Setting pipeline to PLAYING …
New clock: GstSystemClock
Got EOS from element “pipeline0”.
Execution ended after 0:00:00.929473533
Setting pipeline to NULL …
Freeing pipeline …

test.264 output is complete gibberish

fanzh · October 17, 2023, 9:32am

test.264 is H264 data. you can read it by H264 analysis tool, wondering if it can be decoded.

sharan.phadke · October 17, 2023, 9:45am

Yes I am able to play it with vlc media player. Its playing properly

fanzh · October 17, 2023, 10:00am

could you share more logs? please restart the app by this command “export GST_DEBUG=6 && deepstream-app -c xx.txt >1.log 2>1.log”, then test playing RTSP in the docker container. then share the 1.log. you can compress the log by zip.

sharan.phadke · October 17, 2023, 10:57am

1.log (4.0 MB)
log file after running the application

fanzh · October 17, 2023, 2:18pm

sorry, xx.txt is only a sample. please replace xx.txt with your actual configuration file.

sharan.phadke · October 18, 2023, 4:04am

These are the logs after running the application with given config file
log.zip (83.7 MB)

fanzh · October 24, 2023, 7:37am

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

dpsrc gstudpsrc.c:1557:gst_udpsrc_open:e[00m bound, on port 5400
udpsrc gstudpsrc.c:875:gst_udpsrc_fill:e[00m doing select, timeout -1

Sorry for the late reply, Is this still an DeepStream issue to support? Thanks!
from the logs, rtspserver can’t receive data from port 5400 because there is no “gst_udpsrc_fill:^[[00m read packet of 1400 bytes” this kind of printing.

system · November 7, 2023, 7:38am

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.

Topic		Replies	Views
Error streaming RTSP with RTSP sink in deepstream-app DeepStream SDK	8	644	February 7, 2024
DeepStream 5.1 Docker rtsp Unauthorized (401) DeepStream SDK gstreamer	4	1284	November 16, 2021
RTSP stream not work "Stream format not found, dropping the frame" DeepStream SDK rtsp , gstreamer , nvbugs	24	1515	January 3, 2024
Rtsp stream as an input giving generic error DeepStream SDK	6	1377	October 12, 2021
BUG: NVIDIA Deepstream sample Python app not working with RTSP streams DeepStream SDK jetson , deepstream	1	22	October 30, 2024
Deepstream6.4 docker run test1.py issue DeepStream SDK	5	297	April 23, 2024
Running sample deepstream-rtsp-in-rtsp-out but cannot acess RTSP stream DeepStream SDK rtsp , deepstream	9	25	November 21, 2024
Runtime errors from Deepstream 7.1 via GXF on Jetson Xavier NX DeepStream SDK docker , deepstream	28	95	November 21, 2024
Deep stream - Gstreamer NV Plugins DeepStream SDK	18	556	September 11, 2023
How can I get my output address like rtsp or file address DeepStream SDK	13	720	June 16, 2022

Rtsp Sink not working inside docker container

Related topics