LoggerImpl Handler: Error creating native log file.

Describe the bug When starting up my .NET 5 project, I receive the following error.

LoggerImpl Handler: Error creating native log file.

To Reproduce Steps to reproduce the behavior:

1. Created a brand new .NET 5 Console Application
2. Added dockerfile with the following datadog information per the documentation.

RUN curl -LO https://github.com/DataDog/dd-trace-dotnet/releases/download/v1.28.2/datadog-dotnet-apm_1.28.2_amd64.deb
RUN dpkg -i ./datadog-dotnet-apm_1.28.2_amd64.deb

ENV CORECLR_ENABLE_PROFILING=1
ENV CORECLR_PROFILER={846F5F1C-F9AE-4B07-969E-05C26BC060D8}
ENV CORECLR_PROFILER_PATH=/opt/datadog/Datadog.Trace.ClrProfiler.Native.so
ENV DD_INTEGRATIONS=/opt/datadog/integrations.json
ENV DD_DOTNET_TRACER_HOME=/opt/datadog

Expected behavior I expected it to start up without a problem.

Runtime environment (please complete the following information):

• Instrumentation mode: Automatic
• Tracer version: 1.28.2
• OS: Linux
• CLR: .NET 5

Additional context I’m running this is a linux docker container using the .NET 5 runtime.

FROM mcr.microsoft.com/dotnet/runtime:5.0 AS base

Workaround If I add the trace log path environment variable and set it to the default datadog trace location, it works fine.

Here’s a link to a github project with the example. https://github.com/jrmccannon/datadog-1.28.2-test

[andrewlock]

Hi @jrmccannon, thanks for reaching out.

I think you’re missing a line in your Dockerfile that runs the script (included in the deb file) that creates the log path and updates permissions (as described in the docs). For example, you should have something like:

RUN curl -LO https://github.com/DataDog/dd-trace-dotnet/releases/download/v1.28.2/datadog-dotnet-apm_1.28.2_amd64.deb
RUN dpkg -i ./datadog-dotnet-apm_1.28.2_amd64.deb

# Add this line 
RUN chmod +x /opt/datadog/createLogPath.sh && /opt/datadog/createLogPath.sh

# ... other variables as before

This creates the default log directory (/var/log/datadog/dotnet) and adds permissions to write to the directory, alternatively, you can set the log path to an existing directory using DD_TRACE_LOG_DIRECTORY as you mentioned.

The behaviour changed in this PR to print to stderr when we can’t write logs. We will revert that behaviour so that we don’t print to stderr on startup, but you will still want to create the log directory nevertheless 🙂

[andrewlock]

Hi @lucianoeger, yes, the solution will be the same - either create the default log path at /var/log/datadog/dotnet and ensure it is writeable (the easiest way is by running the createLogPath.sh included in the linux packages) or set a different (existing) directory using the DD_TRACE_LOG_DIRECTORY environment variable. We have merged a fix for the issue, which should be included in the next release, but we advise you take the above steps regardless 🙂