Native TritonServer doesn't work on Orin Nano

[pierantonio.merlino]

Hi,

I tried to start the TritonServer binary on my Orin Nano, but I got this errors:

./tritonserver: /lib/aarch64-linux-gnu/libc.so.6: version GLIBC_2.36' not found (required by ./tritonserver) ./tritonserver: /lib/aarch64-linux-gnu/libc.so.6: version GLIBC_2.38’ not found (required by ./tritonserver)
./tritonserver: /lib/aarch64-linux-gnu/libstdc++.so.6: version GLIBCXX_3.4.32' not found (required by ./tritonserver) ./tritonserver: /lib/aarch64-linux-gnu/libc.so.6: version GLIBC_2.38’ not found (required by /home/jetson/Downloads/tritonserver/bin/…/lib/libtritonserver.so)
./tritonserver: /lib/aarch64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.32’ not found (required by /home/jetson/Downloads/tritonserver/bin/…/lib/libtritonserver.so)

It seems that the required libraries are not present on the system with the correct version. I’m using the last Jetpack 6.2 and I downloaded the binaries from here. The release is “Release 2.54.0 corresponding to NGC container 25.01”

Is the native TritonServer supported on this device? Or is there a newer version of the OS with the required libc libraries? Jetpack 6.2 is based on Ubuntu 22.04 that includes libc-2.35, but the executable needes at least libc-2.38 that is provided by Ubuntu 24.04.

Thanks

[AastaLLL]

Hi,

Thanks for updating this.
We will give it a try and provide more info to you later.

[AastaLLL]

Hi,

Thanks for your patience.

The release 2.54.0 iGPU release contains below dependencies:

• This release supports TensorFlow 2.17.0, TensorRT 10.8.0.40, Onnx Runtime 1.20.1, PyTorch 2.6.0a0+ecf3bae40a, Python 3.12 and as well as ensembles.

Different from the libraries that are included in the JetPack 6.2.
To be compatible with JetPack 6.2, please try the 2.49.0 release instead:

• This release supports TensorFlow 2.16.1, TensorRT 10.3.0.26, Onnx Runtime 1.19.0, PyTorch 2.5.0a0+872d972, Python 3.10 and as well as ensembles.

Thanks.

[pierantonio.merlino]

Thanks for the support.