docs: add troubleshooting section

README.md

@@ -218,6 +218,7 @@ designed for modularity and flexibility
 ### Additional Documentation
 
 - [FAQ](docs/user_guide/faq.md)
+- [Troubleshooting Guide](docs/user_guide/troubleshooting.md)
 - [User Guide](docs/README.md#user-guide)
 - [Customization Guide](docs/README.md#customization-guide)
 - [Release Notes](https://docs.nvidia.com/deeplearning/triton-inference-server/release-notes/index.html)

docs/user_guide/troubleshooting.md

@@ -0,0 +1,30 @@
+# Troubleshooting Guide
+
+This guide helps resolve common issues encountered when running Triton Inference Server.
+
+## Common Issues
+
+### Server fails to start
+
+- **Check model repository**: Ensure all models are properly placed in the model repository directory.
+- **Validate model configuration**: Run model validator to check for configuration errors.
+- **Check port availability**: Ensure the HTTP/REST and gRPC ports are not already in use.
+
+### Model fails to load
+
+- **Check model files**: Verify all model files exist and are readable.
+- **Backend compatibility**: Ensure the model format is supported by an available backend.
+- **Configuration errors**: Review the model configuration for syntax errors or invalid parameters.
+
+### Inference requests fail
+
+- **Check client compatibility**: Ensure the client is using the correct inference protocol version.
+- **Input/output mismatch**: Verify input/output tensor names and shapes match model configuration.
+- **Resource exhaustion**: Check GPU memory and ensure the model fits in available memory.
+
+## Getting Help
+
+If issues persist, please:
+- Review the [FAQ](faq.md)
+- Search [existing issues](https://github.com/triton-inference-server/server/issues)
+- Start a [discussion](https://github.com/triton-inference-server/server/discussions)