I use DS4.0.1 on the TX2.
Problem:
I have playback data (jpg-images), that I replay by pushing them to the RTSP-Server. The Deepstream pipeline gets the data via RTSP input source. Here is my config.
[application]
enable-perf-measurement=1
perf-measurement-interval-sec=5
#gie-kitti-output-dir=streamscl
[tiled-display]
enable=0
rows=2
columns=2
width=1280
height=720
gpu-id=0
nvbuf-memory-type=0
[source0]
enable=1
type=4
uri=rtsp://playback:5541/sensor0
num-sources=1
gpu-id=0
nvbuf-memory-type=0
[source1]
enable=1
type=4
uri=rtsp://playback:5541/sensor1
num-sources=1
gpu-id=0
nvbuf-memory-type=0
[source2]
enable=1
type=4
uri=rtsp://playback:5541/sensor2
num-sources=1
gpu-id=0
nvbuf-memory-type=0
[sink1]
enable=1
type=3
container=1
codec=3
sync=0
bitrate=2000000
output-file=/home/nvidia/samples/out.mp4
source-id=0
[osd]
enable=1
gpu-id=0
border-width=1
text-size=15
text-color=1;1;1;1;
text-bg-color=0.3;0.3;0.3;1
font=Arial
show-clock=0
clock-x-offset=800
clock-y-offset=820
clock-text-size=12
clock-color=1;0;0;0
nvbuf-memory-type=0
[streammux]
gpu-id=0
live-source=1
batch-size=3
batched-push-timeout=40000
width=608
height=608
enable-padding=0
nvbuf-memory-type=0
[primary-gie]
enable=1
gpu-id=0
batch-size=3
bbox-border-color0=1;0;0;1
bbox-border-color1=0;1;1;1
bbox-border-color2=0;1;1;1
bbox-border-color3=0;1;0;1
nvbuf-memory-type=0
interval=0
gie-unique-id=1
model-engine-file=model_b3_fp16.engine
config-file=yolo_config.txt
It happens very regularily (not always), that I get a message “reference in DPB was never decoded” (message occurs only once in the beginning). In this case one of the three streams drops input frames and starts to push later frames to the streammux. This results that the streammux-batch contains image frames from completely different timestamps. If I get the message two times, two streams are off.
I observed some more regularities. When I push frames to the RTSP server on 1fps, the “dropping stream” is always 10 frames before the correct streams. When I push frames to the RTSP server on 2fps, the “dropping stream” is always 20 frames before the correct streams. So there seems to be a factor of 10??
“reference in DPB was never decoded” also happens with only one RTSP source.
I also tried to disable the tiled-display. It didn’t make a difference.
Here an example output of the pipeline running (two streams are off in this example)
Deserialize yoloLayerV3 plugin: yolo_17
Deserialize yoloLayerV3 plugin: yolo_24
Deserialize yoloLayerV3 plugin: yolo_31
Runtime commands:
h: Print this help
q: Quit
p: Pause
r: Resume
**PERF: FPS 0 (Avg) FPS 1 (Avg) FPS 2 (Avg)
**PERF: 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)
** INFO: <bus_callback:189>: Pipeline ready
Opening in BLOCKING MODE
NvMMLiteOpen : Block : BlockType = 261
NVMEDIA: Reading vendor.tegra.display-size : status: 6
NvMMLiteBlockCreate : Block : BlockType = 261
Creating LL OSD context new
Opening in BLOCKING MODE
NvMMLiteOpen : Block : BlockType = 261
NVMEDIA: Reading vendor.tegra.display-size : status: 6
NvMMLiteBlockCreate : Block : BlockType = 261
Opening in BLOCKING MODE
NvMMLiteOpen : Block : BlockType = 261
NVMEDIA: Reading vendor.tegra.display-size : status: 6
NvMMLiteBlockCreate : Block : BlockType = 261
** INFO: <bus_callback:175>: Pipeline running
**PERF: 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)
**PERF: 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)
reference in DPB was never decoded
reference in DPB was never decoded
**PERF: 1.00 (1.00) 1.00 (1.00) 1.00 (1.00)
**PERF: 1.00 (1.00) 1.00 (1.00) 1.00 (1.00)
**PERF: 1.00 (1.00) 1.00 (1.00) 1.00 (1.00)
**PERF: 1.00 (1.00) 1.00 (1.00) 1.00 (1.00)
**PERF: 1.00 (1.00) 1.00 (1.00) 1.00 (1.00)
If I run the same code on GPU “Quadro P2000”, I don’t get these problems. Everything runs fine.
Thanks.