Skip to content

Latest commit

 

History

History
380 lines (267 loc) · 13 KB

editors.md

File metadata and controls

380 lines (267 loc) · 13 KB
Raw

Editors and IDEs

There are several primary ways you can connect an IDE to your Coder workspace:

  1. VS Code remote SSH with local VS Code
  2. VS Code in the browser with code-server
  3. JetBrains Gateway and SSH
  4. Jupyter Notebook
  5. JupyterLab
  6. RStudio
  7. Any local editor with 1-way file synchronization or 2-way file synchronization over SSH

VS Code remote SSH

Once you've set up SSH access to Coder, you can work on projects from your local VS Code, connected to your Coder workspace for compute, etc.

  1. Open VS Code locally.
  2. Make sure that you've installed Remote - SSH extension
  3. In VS Code's left-hand nav bar, click Remote Explorer and right-click on a workspace to connect

If Coder is deployed air-gapped (no Internet), you need to configure your VS Code's setting with remote.SSH.allowLocalServerDownload enabled so the extension will install the VS Code Server on the client first and then copy it over to the Coder workspace via SCP.

For further troubleshooting steps, see Troubleshooting hanging or failing connections

VS Code Remote Explorer

VS Code in the browser

Launch VS Code in the browser from the workspaces page by clicking the Code Web icon.

Launch a workspace

Code Web is Coder's open-source project code-server.

Opening files via the terminal

You can open files from your Coder workspace in VS Code via the terminal. We recommend creating an alias to the underlying code-server executable so that you can use the command code for this process:

alias code="/var/tmp/coder/code-server/bin/code-server -r"

Then, to open a file (e.g., personalize.log):

code personalize.log

Alternatively, if you would like to use just the code-server executable, add it to your PATH:

export PATH=$PATH:/var/tmp/coder/code-server/bin

Then, to open a file (e.g., personalize.log):

code-server -r personalize.log

If you're using Coder's web terminal, make sure that you've opened a Code Web session. If, however, you're using the web IDE's terminal, the file contents will appear in the IDE.

JetBrains Gateway with SSH

Gateway is JetBrains' remote development solution. JetBrains has suspended Projector (the browser-based option) therefore Coder no longer provides examples or support.

By default, Gateway will download the IDE from jetbrains.com into the Coder workspace during the setup. If you are air-gapped or want to leverage a JetBrains IDE in your workspace for faster setup, you can point to an already-installed JetBrains IDE. See the configuration at the end of this Gateway section.

Requirements:

  • SSH access to Coder must already be configured
  • Your Coder workspace must be running. Gateway needs compute resources, so monitor your resource usage on the Coder dashboard and adjust accordingly.
  • If you use a premium JetBrains IDE (e.g., GoLand, IntelliJ IDEA Ultimate), you will still need a license to use it remotely with Coder.

  1. Download and install JetBrains Toolbox. Locate JetBrains Gateway in the Toolbox list and click Install.

    JetBrains Toolbox

  2. Open JetBrains Gateway and click Connect via SSH within the Run the IDE Remotely section.

    Open JetBrains Gateway

  3. Click the small gear icon to the right of the Connection field, then the + button on the next screen to create a new configuration.

    Connect Gateway to SSH

  4. Enter your Coder workspace host name in Host (e.g., coder.mark-intellij), 22 in Port, coder in User name, and change Authentication Type to OpenSSH config and authentication agent. You can find the workspace host names in ~/.ssh/config. Leave the local port field blank. Click Test Connection.

    Gateway SSH Configurations

  5. Choose your new connection from the drop-down and click Check Connection and Continue

    Connect to SSH

  6. The default behavior is to select a JetBrains IDE from the IDE version drop-down and download it from jetbrain.com. Choose the IDE installed in your Coder workspace, and click the folder icon and select your /home/coder directory in your Coder workspace.

    Select JetBrains IDE and working directory

    If you ran remote-dev-server.sh (see note below) before starting the config setup, JetBrains will detect your already installed IDE in the drop-down.

    Select JetBrains IDE and working directory

  7. Gateway will open the JetBrains client connected to the remotely installed IDE.

    A running JetBrains IDE in Gateway

Using an existing JetBrains installation in the workspace

If you would like to use an existing JetBrains IDE in a Coder workspace (or you are air-gapped, and cannot reach jetbrains.com), run the following script in the JetBrains IDE directory to point the default Gateway directory to the IDE directory. This step must be done before configuring Gateway.

cd /opt/idea/bin
./remote-dev-server.sh registerBackendLocationForGateway

Here is the JetBrains article explaining this IDE specification.

Alternative SSH key algorithms and Gateway

If your Coder deployment is configured with ECDSA ssh key algorithm, change the Gateway authentication type to Key pair and create the Coder public ssh key in your local ~/.ssh directory with ssh-keygen -y -f:

~/.ssh/coder_enterprise | tee ~/.ssh/coder_enterprise.pub

Support & troubleshooting

This article outlines troubleshooting steps with Gateway. JetBrains product support including their Issue Trackers are here.

Jupyter Notebook

Jupyter Notebook is the original web IDE for creating Notebooks used in data science, machine learning and analytics projects. By default, any Coder workspace with the Jupyter project installed (in /usr/local/bin/jupyter) will render the icon to launch Jupyter Notebook.

Jupyter Notework

To use Jupyter Notebook in a Coder workspace, build a Dockerfile with Jupyter project installed as shown below:

# Dockerfile to install Jupyter Notebook
FROM codercom/enterprise-base:ubuntu

USER root

RUN pip3 install jupyter notebook

USER coder

JupyterLab

JupyterLab is the next-generation web-based IDE for data science and Python using documents called Notebooks.

JupyterLab

There are three methods to install and access JupyterLab in Coder. All require JupyterLab to be installed in the Dockerfile via pip3 install jupyterlab.

The first method renames the jupyter binary and copies a new jupyter that adjusts the arguments passed to the jupyter binary to tell Coder to launch JupyterLab instead of Notebook.

FROM codercom/enterprise-base:ubuntu

USER root

RUN pip3 install jupyterlab
RUN pip3 install jupyter notebook

RUN mv /usr/local/bin/jupyter /usr/local/bin/jupyter.py

COPY jupyter /usr/local/bin/jupyter

RUN chmod +x jupyter

USER coder

Below is an example jupyter script with the lab arguments. This file must be located in the same directory as the Dockerfile to be copied during docker build

#!/bin/bash
# Replace all "NotebookApp" settings with ServerApp settings.
args=${@//LabApp/"ServerApp"}
# Replace 'notebook' with 'lab' to launch juypter lab
args=${args/notebook/"lab"}

jupyter.py ${args}

The second method to run JupyterLab is with a dev URL and launching JupyterLab via supervisord in the configure script. The benefit of this approach is it is completely independent of Coder's IDE launching mechanism and relies only on a generic dev URL.

JupyterLab as a dev URL

FROM codercom/enterprise-base:ubuntu

USER root

RUN pip3 install jupyterlab
RUN pip3 install jupyter notebook

# configure script to create a dev URL and launch JupyterLab
COPY ["configure", "/coder/configure"]
RUN chmod +x /coder/configure

# install supervisord
RUN apt-get update && apt-get install -y supervisor
RUN mkdir -p /var/log/supervisor
COPY supervisord.conf /etc/supervisor/conf.d/supervisord.conf

# change back to the coder user
USER coder

The configure script installs supervisord

#!/bin/bash

echo 'create dev URL for JupyterLab'
coder urls create $CODER_WORKSPACE_NAME 8888 --name jupyterlab

echo 'start supervisord and JupyterLab'
sudo /usr/bin/supervisord

The supervisord.conf launches JupyterLab. This file must be located in the same directory as the Dockerfile to be copied during docker build

[supervisord]
nodaemon=false
environment=HOME=/home/coder

[program:jupyterlab]
command=/usr/local/bin/jupyter lab --ip='*' --NotebookApp.token='' --NotebookApp.password=''
user=coder
directory=/home/coder

The third method to access JupyterLab is locally using the SSH port forward command: ssh -L 8888:localhost:8888 coder.jupyterlab. Alternatively, you can use the Coder CLI to port forward using: coder tunnel jupyterlab 8888 8888. Now, open a local browser and navigate to https://localhost:8888.

RStudio

Coder supports RStudio. To create a workspace that lets you use RStudio:

  1. Create a custom image with RStudio installed, rserver in PATH.

    To do this, you can refer to the sample Dockerfile below, which installs RStudio Server Open Source to log in with username coder and password rstudio.

    This Dockerfile approach works now with latest versions of RStudio including 2022-2-1.

    FROM codercom/enterprise-base:ubuntu

USER root

# Install dependencies
RUN apt-get update && \
DEBIAN_FRONTEND="noninteractive" apt-get install --yes \
r-base \
gdebi-core

# Install RStudio
RUN wget https://download2.rstudio.org/server/bionic/amd64/rstudio-server-2022.02.1-461-amd64.deb && \
gdebi --non-interactive rstudio-server-2022.02.1-461-amd64.deb

# Ensure rstudio files can be written to by the coder user.
RUN chown -R coder:coder /var/lib/rstudio-server
RUN echo "server-pid-file=/tmp/rstudio-server.pid" >> /etc/rstudio/rserver.conf
RUN echo "server-data-dir=/tmp/rstudio" >> /etc/rstudio/rserver.conf
RUN echo "www-frame-origin=same" >> /etc/rstudio/rserver.conf
RUN echo "server-user=coder" >> /etc/rstudio/rserver.conf

# Remove the following line if you do not run Coder on https
RUN echo "server-add-header=X-Forwarded-Proto: https" >> /etc/rstudio/rserver.conf

# Assign password "rstudio" to coder user.
RUN echo 'coder:rstudio' | chpasswd

# Assign locale
RUN locale-gen en_US.UTF-8

# Run as coder user
USER coder

# Add RStudio to path
ENV PATH /usr/lib/rstudio-server/bin:${PATH}

  2. Create a workspace using the image you created in the previous step.

  3. At this point, you can go to Applications to launch RStudio.

    Applications with RStudio launcher

    Coder auto-signs in using the Unix user (whose username and password you defined in your custom image above).

    RStudio may take a few additional seconds to start launch after the workspace is built.

    All RStudio data is stored in the home directory associated with the user you sign in as, since this ensures that your data is saved if Coder shuts down or rebuilds your environment.

Logging

You can find your IDE logs in the following places:

  • For code-server: ~/.local/share/code-server/logs/
  • For JetBrains IDEs: .cache/JetBrains/<JetBrains-IDE>/log/<IDE>.log