DataDog
[BUG] Jetson integration does not handle Jetson Orin
Agent Environment Machine: NVIDIA Jetson Orin NX 16GB
# kubectl exec -it -n monitoring datadog-2h7wv -- agent version
Defaulted container "agent" out of: agent, trace-agent, process-agent, init-volume (init), init-config (init)
Agent 7.45.0 - Commit: 964e770 - Serialization version: v5.0.81 - Go version: go1.19.9
# cat /etc/nv_tegra_release
# R35 (release), REVISION: 3.1, GCID: 32827747, BOARD: t186ref, EABI: aarch64, DATE: Sun Mar 19 15:19:21 UTC 2023
Describe what happened:
The Jetson check fails with could not parse voltage fields from this code since the regex cannot parse the output of tegrastats.
We are using the latest Jetson model: Orin. The results of tegrastats on this device return in a different format to previous, and the Jetson integration cannot handle them:
~# tegrastats
06-16-2023 10:45:09 RAM 6334/15388MB (lfb 1770x4MB) SWAP 491/7694MB (cached 0MB) CPU [6%@729,9%@729,5%@729,16%@729,off,off,off,off] EMC_FREQ 0%@2133 GR3D_FREQ 0%@611 GR3D2_FREQ 0%@0 VIC_FREQ 729 APE 174 [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] VDD_IN 5299mW/5299mW VDD_CPU_GPU_CV 773mW/773mW VDD_SOC 1424mW/1424mW
06-16-2023 10:45:10 RAM 6335/15388MB (lfb 1770x4MB) SWAP 491/7694MB (cached 0MB) CPU [16%@1344,11%@1344,12%@1331,14%@1497,off,off,off,off] EMC_FREQ 0%@2133 GR3D_FREQ 0%@611 GR3D2_FREQ 0%@0 VIC_FREQ 729 APE 174 [email protected] [email protected] [email protected] [email protected] [email protected] GPU@46C [email protected] [email protected] [email protected] VDD_IN 5218mW/5258mW VDD_CPU_GPU_CV 691mW/732mW VDD_SOC 1424mW/1424mW
Describe what you expected: The Jetson check should work on a Jetson Orin with the latest JetPack.
Steps to reproduce the issue: Enable the Jetson integration on a Jetson Orin.
Additional environment details (Operating System, Cloud provider, etc): The test for the Jetson integration lists the different models of Jetson that it handles.
I've provided an example from the Orin above. If the test is able to work with that then that's all that's needed.
