6. Troubleshooting

In this section you’ll learn:

Overview

As a Posit Workbench Administrator you may get reports of issues that are:

  • specific to a user’s R or Python code
  • Server resource issues
  • Software configuration issues
  • Networking issues

In this section we’ll provide some general guidelines that can help you focus in on the cause of issues. In the lab component of this section you’ll get experience using the logs to actually diagnose issues.

What is the scope of the issue?

Understanding where this issue is and is not occurring is one of the fastest ways to hone in on the problem. For example, errors may originate from:

  • User code

  • an installed package

  • the programming language

  • the IDE

  • Workbench

  • the Linux OS

Some questions you’ll want to consider:

  • Does it affect one user, the entire server, or the whole cluster?
  • Can you reproduce the problem?
  • Is the error reproducible from the command line outside of Workbench?
  • Identify the root cause of the error

    • Any recent changes?

    • Specific users or groups involved?

    • Specific code being run?

Reviewing the logs

As you narrow in on the scope of the issue it can be useful to review the server logs. Below are some tips on navigating and interpreting the relevant logs.

  • For a broad starting point, you can check the status of rstudio-server and rstudio-launcher, this will let you know if both of these services are running and will show some of the recently logged messages for each service.

  • Review the default logs and file locations from the Administrator’s Guide. Another broad starting point for troubleshooting can be to check the Posit Workbench Monitor, which collects errors and warnings from multiple executables.

  • For issues inside of successfully started? R sessions, you should review the logs for rsession and rworkspaces

  • For issues starting sessions in environments where Launcher is enabled (i.e. you have Jupyter or VS Code installed, or have integrated with Kubernetes or Slurm), you should review the logs for the Job Launcher, both the main process and for the specific plugin that you are using (i.e., Kubernetes, Slurm, or Local).

  • For most other issues, you’ll also want to review the rstudio-server logs.

  • Workbench also integrates with a number of Linux processes, so logs from some of these processes (e.g. PAM, sssd) can also be important to review depending on the nature of the issue you are experiencing.

  • If you are still unable to diagnose an issue after reviewing the logs, you should enable debug logging.

Filing a support ticket

By purchasing Posit software you have access to Posit Support, which is outlined in our Support Agreement.

You have access to an automated ticket and reporting service through which support requests are logged and ticket ID(s) issued within thirty minutes. This service can be reached at support.posit.co or a direct email can be sent to support@posit.co.

When writing your ticket it is helpful if you share the scope of the issue, the troubleshooting steps you’ve already taken, and the server diagnostics. You can create a diagnostic report to attach to the ticket by running: sudo rstudio-server run-diagnostic.

Troubleshooting Lab

🚀 Launch the exercise environment!

In the exercise environment you will get experience:

  • finding and reading the Workbench logs

  • using logs to diagnose problems in the environment

Go to: 7. Scaling

Back to top