•Hardware Platform : Jetson •DeepStream Version 5.0 •JetPack Version 4.4 [L4T 32.4.3] •TensorRT Version 7.1.3 •Issue Type(question/bugs)
Using deepstream-test5 with yolov3 that is bundled with DS over 5 RTSP cameras is throwing a segmentation fault after pipeline ran for some time. It goes something like:
*** DeepStream: Launched RTSP Streaming at rtsp://localhost:8555/ds-test ***
Opening in BLOCKING MODE
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-5.0/lib/libnvds_nvdcf.so
gstnvtracker: Batch processing is ON
gstnvtracker: Past frame output is OFF
[NvDCF] Initialized
WARNING: [TRT]: Using an engine plan file across different models of devices is not recommended and is likely to affect performance or even cause errors.
Deserialize yoloLayerV3 plugin: yolo_83
Deserialize yoloLayerV3 plugin: yolo_95
Deserialize yoloLayerV3 plugin: yolo_107
INFO: [Implicit Engine Info]: layers num: 4
0 INPUT kFLOAT data 3x608x608
1 OUTPUT kFLOAT yolo_83 255x19x19
2 OUTPUT kFLOAT yolo_95 255x38x38
3 OUTPUT kFLOAT yolo_107 255x76x76
Runtime commands:
h: Print this help
q: Quit
p: Pause
r: Resume
NOTE: To expand a source in the 2D tiled display and view object details, left-click on the source.
To go back to the tiled display, right-click anywhere on the window.
** INFO: <bus_callback:181>: Pipeline ready
Thu Apr 8 03:40:21 2021
**PERF: 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)
Opening in BLOCKING MODE
Opening in BLOCKING MODE
Opening in BLOCKING MODE
Opening in BLOCKING MODE
Opening in BLOCKING MODE
** INFO: <bus_callback:167>: Pipeline running
WARNING; playback mode used with URI [rtsp://****/cam0_0] not conforming to timestamp format; check README; using system-time
WARNING; playback mode used with URI [rtsp://****/cam0_0] not conforming to timestamp format; check README; using system-time
WARNING; playback mode used with URI [rtsp://****/cam0_0] not conforming to timestamp format; check README; using system-time
WARNING; playback mode used with URI [rtsp://****/cam0_0] not conforming to timestamp format; check README; using system-time
Thu Apr 8 03:40:26 2021
**PERF: 9.56 (9.29) 10.38 (10.22) 10.07 (9.80) 10.06 (9.78) 9.26 (9.09)
.
After few hours
.
Thu Apr 8 04:26:31 2021
**PERF: 10.11 (9.69) 10.21 (10.08) 10.28 (10.02) 10.11 (9.90) 10.37 (9.69)
Quitting
Segmentation fault
Where did you got the yolov3 models? Are you running the case as described in /opt/nvidia/deepstream/deepstream-5.1/sources/objectDetector_Yolo/README?
config_infer_primary_yoloV3.txt (3.4 KB) test.txt (8.1 KB)
Here are the files. Here’s some more info: I tried clearing gstreamer cache. There are other processes running on the AGX too: 5 python processes, kafka broker and mongod. Mongo’s dbpath is set to an external SDcard but its mounted inside the deepstream-test5 folder.
There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks
I have run the case for 2 hours. It works well. Can you provide a stable reproduce method?
We switched to a new AGX and the problem was solved. I’m guessing it was due to rtsp connection issue from the source with the previous device because recently I ran into a test video(mp4) that gave segmentation fault. So, it should be the consequence of bad data that was produced in earlier. Thank you for your support, you do great work.