Skip to content

Return to Table of Contents

Restarting the OpenSARLab Server and Notebook Kernel

Restarting the OpenSARLab Server

Overview

Restarting the server triggers the nbgitpuller.

Consider a case where:

  • You have deleted or altered a notebook in the ASF notebook library and want to retrieve the original.
  • There is a notebook update that you wish to pull in changes from the asf-jupyter-notebook repo.

A quick solution in either of those cases is to restart your server to run the nbgitpuller.

NB: If you are comfortable with git, you could do a git pull from the terminal or in a notebook instead. However, do not push your changes as it may interfere with the nbgitpuller

Steps to Restart the Server

  1. Select Hub Control Panel from the File menu

    Select Hub Control Panel from the File menu

    or

    Click the Shutdown and Logout Page button in the upper right corner of the screen.

    Click the Shutdown and Logout Page button in the upper right corner of the screen

  2. Click the Stop My Server Button

    Click the Stop My Server button

  3. Click the Start My Server Button, which may take a few seconds to appear.

    Click the Start My Server button

  4. Click The Launch Server Button

    Click the Launch Server button

  5. Select a Server Profile and Click the Start button

    Select a server option and click the start button

  6. Wait for the server to start; this may take some time.

    Wait for the server to start

  7. Optional: Click The Event Log arrow for detailed startup status information. You may use this information to send a report to the admin if your server does not start.

    Click the event log arrow

Changing a Notebook Kernel

Overview

Notebooks in OpenScienceLab run in a variety of conda environments. Since each kernel has different packages, you must pick the right one to confirm it has the required packages.

How to Switch Kernel

  1. Click the upper right corner. It should have the name of the kernel that you are currently using.

  2. A window with a list of built kernel will pop up. Select whichever kernel you need.

    Select Kernel

NB: To build a kernel, refer to the conda environments section.

Restarting a Jupyter Notebook Kernel

Overview

Variables and their assigned values are stored in memory as you run code cells in a notebook. As a result, rerunning a previously ran notebook without restarting the kernel may trigger some issues.

Uncleared variables are problematic for various reasons, such as:

  • Increase the file size of the notebook.
  • They occupy limited memory resources.
  • *Previously defined values may cause unexpected results during the rerun ().

NB_: _Refer to the Rerunning a Notebook section in the How to Run a Jupyter Notebook* section.

The solution for this is to restart the kernel to clear notebook data.

How to Clear Notebook

Select one of the following from the Kernel Menu:

  • Restart Kernel: Restarts the kernel but leave the old code cell output in place.
  • Restart Kernel and Clear All Outputs...: Restarts the kernel and removes old code cell output. This is generally the preferred option.
  • Restart Kernel and Run up to Selected Cell...: Restarts the kernel and run up to the cell where you selected.

  • Restart Kernel and Run All Cells...: Restarts the kernel and runs all the code cells. This only works if the notebook does not require input from the user.

    Select restart and clear all from the kernel menu

    For most use cases, select Restart Kernel and Clear All Outputs...