## Summary Nests the CLI reference by subcommand so the Command Line sidebar collapses from 156 flat entries to 49 top-level verbs. Each page's URL follows the invocation, its page title is the full command as inline code, and its sidebar label is only its own verb as inline code. | | Before | After | | --- | --- | --- | | File | `docs/reference/cli/oauth2-provider_dcr_disable.md` | `docs/reference/cli/oauth2-provider/dcr/disable.md` | | URL | `/reference/cli/oauth2-provider_dcr_disable` | `/reference/cli/oauth2-provider/dcr/disable` | | Page title (front matter) | `oauth2-provider dcr disable` | `` `coder oauth2-provider dcr disable` `` | | Sidebar label (manifest) | `oauth2-provider dcr disable` | `` `disable` `` | Resolves DOCS-140 (coder/coder half). Also fixes the Makefile part of DOCS-1073 / DOCS-912, which this change depends on. ## Changes - `scripts/clidocgen`: pages nest by command path. A command with visible subcommands becomes its directory's `index.md`, so `oauth2-provider/index.md` and `oauth2-provider/dcr/index.md` sit above `oauth2-provider/dcr/disable.md`. Subcommand table links are relative to the parent page. - `scripts/clidocgen`: the manifest "Command Line" section is built from the command tree (nested, sorted by name, hidden commands skipped) instead of a flat list. The redundant `coder` child that pointed back at `reference/cli/index.md` is gone. - `Makefile`: the CLI rule now replaces `docs/reference/cli/` wholesale, so stale pages get removed. The manifest rule now runs `clidocgen -manifest-only` on top of the API staging manifest, so the CLI nav is generated instead of edited by hand. As a result, the nine pages that had no nav entry (`ai-gateway*`, `secret enable/disable`, `server fix-oidc-links`) now have one. - `scripts/configdocgen` and 48 prose pages: links point at the new paths (for example `reference/cli/server.md` is now `reference/cli/server/index.md`). ## How to review/QA this PR - `docs/reference/cli/**/*` (approximately 190 pages): Each of these pages are generated automatically. Some pages appear deleted, and some pages appear new. Every deleted page corresponds to a new page somewhere else, so no information is being deleted. The docs in this section should adhere to the following, so a test in the build preview for each of these items should be satisfactory: - [ ] Each subcommand of a command gets nested as a child instead of being displayed as a flat list. - [ ] Each parent command is stored in an `index.md` file. - [ ] The title of each page inside the page should now feature the full `coder XYZ` command. - [ ] The title of each page inside the page should now be formatted as inline code. - [ ] The title of each page in the sidebar should be the most granular subcommand of the whole command string (if the page's title is `coder ai-gateway keys create`, the sidebar label is `create`). - [ ] The title of each page in the sidebar should now be formatted as inline code. - [ ] The page slug for each page is now the most granular subcommand. - [ ] All links now point to the correct page. - `docs/manifest.json`: The changes here reflect the updated structure of the `docs/reference/cli` section. - `docs/**/*.md` outside of `docs/reference/cli/**/*` (approximately 50 pages): The extent of the changes in these pages are just updating the paths to point to the new page locations. - [ ] The links on the page all point to the correct new path in `docs/reference/cli`. - `scripts/clidocgen` (4 files): This is the Go code responsible for generating everything inside of `docs/reference/cli/**/*`. A typical code review here should be satisfactory. - `scripts/configdocgen` (2 files): these two files just have updates to the raw `.md` path used by `docs/admin/setup/configuration-reference.md` and updates the tests to reflect that. - `Makefile`: This updates the `make` targets for generating the CLI docs to fit the new structure. ## Merge ordering Without the renderer PRs deployed, both docs sites show literal backticks in titles and sidebar labels. Without the redirects, the old flat URLs return 404. 1. Deploy before merging this PR: coder/docs-engine#176 (beta-docs renders these titles as inline code) and coder/coder.com#1111 (classic docs renders them as inline code). Both are safe to ship first, because titles without backticks render as before. 2. Merge this PR. 3. Merge coder/coder.com#1112 (redirects from the old flat URLs) only once this content reaches the mainline release branch that serves unversioned `/docs`. ## Testing - `go test ./scripts/clidocgen/ ./scripts/configdocgen/ ./scripts/docshtmlcheck/`: pass. New tests cover page paths, subcommand links, the nested manifest, and front matter quoting. - `make docs/reference/cli/index.md docs/manifest.json docs/admin/setup/configuration-reference.md`, then a forced rerun: the second run produced no diff. - Checked the manifest: all 164 CLI entries resolve to a file, no page is missing a nav entry, and everything outside "Command Line" is byte-identical. - Relative `.md` links across `docs/`: 0 broken, same as `main`. - `make pre-commit` and `make pre-push`: pass. On earlier pre-push attempts, different unrelated vitest files timed out under host load. Those files pass when run alone, and the final push passed with `PARALLEL_JOBS=1`. <details> <summary>Implementation plan and decision log</summary> **Goal (from DOCS-140 and the requester):** URLs follow the command hierarchy. Pages are titled with the full command, and sidebar entries with the last verb only. Both are formatted as inline code. **Contract with the renderers:** a title wrapped entirely in one pair of backticks (``^`[^`]+`$``) renders as inline code. Plain-text consumers (`<title>`, meta, search, llms corpus) strip the backticks. Page title comes from front matter `title`, and the sidebar label from the manifest node `title`. **Decisions** - A parent command is `dir/index.md` rather than a sibling `dir.md`. The root CLI page is already `index.md`, and both docs-engine and the legacy coder.com renderer map `x/index.md` to `/x`. - The manifest is built from the command tree, not from the written files, so ordering is deterministic and the nesting matches the pages exactly. - The manifest discard bug (DOCS-1073 item 1) is fixed here, because without that fix the generated nested manifest would never reach the tree. DOCS-1073 items 2 through 4 (a prose fix and a hidden-command policy) are out of scope. - `-manifest-only` keeps the existing parallel `make gen` layout: the API rule still owns `_gen/manifest-staging.json`, and CLI regeneration doesn't retrigger swag. - The renderer and redirect work lives in the companion PRs in coder/docs-engine and coder/coder.com. It is linked above and not duplicated here. </details> --- > [!NOTE] > Generated by Coder Agents on behalf of @nickvigilante.
21 KiB
title
| title |
|---|
| FAQs |
Frequently asked questions on Coder OSS and licensed deployments. These FAQs come from our community and customers, feel free to contribute to this page.
For other community resources, see our GitHub discussions, or join our Discord server.
How do I add a Premium trial license?
Visit https://coder.com/trial or contact sales@coder.com to get a trial key.
You can add a license through the UI or CLI
Coder UI
-
With an
Owneraccount, go to Admin settings > Deployment. -
Select Licenses from the sidebar, then Add a license:
-
On the Add a license screen, drag your
.jwtlicense file into the Upload Your License section, or paste your license in the Paste Your License text box, then select Upload License:
Coder CLI
-
Ensure you have the Coder CLI installed.
-
Save your license key to disk and make note of the path.
-
Open a terminal.
-
Log in to your Coder deployment:
coder login <access url> -
Run
coder licenses add:-
For a
.jwtlicense file:coder licenses add -f <path to your license key> -
For a text string:
coder licenses add -l 1f5...765
-
Visit the licensing documentation for more information about licenses.
I'm experiencing networking issues, so want to disable Tailscale, STUN, Direct connections and force use of websocket
The primary developer use case is a local IDE connecting over SSH to a Coder workspace.
Coder's networking stack has intelligence to attempt a peer-to-peer or Direct connection between the local IDE and the workspace. However, this requires some additional protocols like UDP and being able to reach a STUN server to echo the IP addresses of the local IDE machine and workspace, for sharing using a Wireguard Coordination Server. By default, Coder assumes Internet and attempts to reach Google's STUN servers to perform this IP echo.
Operators experimenting with Coder may run into networking issues if UDP (which STUN requires) or the STUN servers are unavailable, potentially resulting in lengthy local IDE and SSH connection times as the Coder control plane attempts to establish these direct connections.
Setting the following flags as shown disables this logic to simplify troubleshooting.
| Flag | Value | Meaning |
|---|---|---|
CODER_BLOCK_DIRECT |
true |
Blocks direct connections |
CODER_DERP_SERVER_STUN_ADDRESSES |
"disable" |
Disables STUN |
CODER_DERP_FORCE_WEBSOCKETS |
true |
Forces websockets over Tailscale DERP |
How do I configure NGINX as the reverse proxy in front of Coder?
This tutorial in our docs explains in detail how to configure NGINX with Coder so that our Tailscale Wireguard networking functions properly.
How do I hide some of the default icons in a workspace like VS Code Desktop, Terminal, SSH, Ports?
The visibility of Coder apps is configurable in the template. To change the
default (shows all), add this block inside the
coder_agent
of a template and configure as needed:
display_apps {
vscode = false
vscode_insiders = false
ssh_helper = false
port_forwarding_helper = false
web_terminal = true
}
This example will hide all built-in
coder_app
icons except the web terminal.
I want to allow code-server to be accessible by other users in my deployment
We don't recommend that you share a web IDE, but if you need to, the following deployment environment variable settings are required.
Set deployment (Kubernetes) to allow path app sharing:
# allow authenticated users to access path-based workspace apps
- name: CODER_DANGEROUS_ALLOW_PATH_APP_SHARING
value: "true"
# allow Coder owner roles to access path-based workspace apps
- name: CODER_DANGEROUS_ALLOW_PATH_APP_SITE_OWNER_ACCESS
value: "true"
In the template, set
coder_app
share
option to authenticated and when a workspace is built with this template, the
pretty globe shows up next to path-based code-server:
resource "coder_app" "code-server" {
...
share = "authenticated"
...
}
I installed Coder and created a workspace but the icons do not load
An important concept to understand is that Coder creates workspaces which have an agent that must be able to reach the control plane.
If the CODER_ACCESS_URL is not
accessible from a workspace, the workspace may build, but the agent cannot reach
Coder, and thus the missing icons. e.g., Terminal, IDEs, Apps.
By default, coder server automatically creates an Internet-accessible
reverse proxy so that workspaces you create can reach the control plane.
If you are doing a standalone install, e.g., on a MacBook and want to build workspaces in Docker Desktop, everything is self-contained and workspaces (containers in Docker Desktop) can reach the control plane.
coder server --access-url http://localhost:3000 --address 0.0.0.0:3000
Even coder server which creates a reverse proxy, will let you use
http://localhost to access Coder from a browser.
I updated a template, and an existing workspace based on that template fails to start
When updating a template, be aware of potential issues with input variables. For example, if a template prompts users to choose options like a code-server VS Code IDE release, a container image, or a VS Code extension, removing any of these values can lead to existing workspaces failing to start. This issue occurs because the Terraform state will not be in sync with the new template.
However, a lesser-known CLI sub-command,
coder update, can resolve this issue. This
command re-prompts users to re-enter the input variables, potentially saving the
workspace from a failed status.
coder update --always-prompt <workspace name>
I'm running coder on a VM with systemd but latest release installed isn't showing up
Take, for example, a Coder deployment on a VM with a 2 shared vCPU systemd
service. In this scenario, it's necessary to reload the daemon and then restart
the Coder service. This prevents the systemd daemon from trying to reference
the previous Coder release service since the unit file has changed.
The following commands can be used to update Coder and refresh the service:
curl -fsSL https://coder.com/install.sh | sh
sudo systemctl daemon-reload
sudo systemctl restart coder.service
I'm using the built-in Postgres database and forgot admin email I set up
- Run the
coder servercommand below to retrieve thepsqlconnection URL which includes the database user and password. psqlinto Postgres, and do a select query on theuserstable.- Restart the control plane, pull up the Coder UI and log in (you will still need your password)
coder server postgres-builtin-url
psql "postgres://coder@localhost:53737/coder?sslmode=disable&password=I2S...pTk"
How to find out Coder's latest Terraform provider version?
Coder is on the HashiCorp's Terraform registry. Check this frequently to make sure you are on the latest version.
Sometimes, the version may change and resource configurations will either
become deprecated or new ones will be added when you get warnings or errors
creating and pushing templates.
How can I set up TLS for my deployment and not create a signed certificate?
Caddy is an easy-to-configure reverse proxy that also automatically creates
certificates from Let's Encrypt.
Install docs here You
can start Caddy as a systemd service.
The Caddyfile configuration will appear like this where 127.0.0.1:3000 is your
CODER_ACCESS_URL:
coder.example.com {
reverse_proxy 127.0.0.1:3000
tls {
issuer acme {
email user@example.com
}
}
}
I'm using Caddy as my reverse proxy in front of Coder. How do I set up a wildcard domain for port forwarding?
Caddy requires your DNS provider's credentials to create wildcard certificates. This involves building the Caddy binary from source with the DNS provider plugin added. e.g., Google Cloud DNS provider here
To compile Caddy, the host running Coder requires Go. Once installed, replace
the existing Caddy binary in usr/bin and restart the Caddy service.
The updated Caddyfile configuration will look like this:
*.coder.example.com, coder.example.com {
reverse_proxy 127.0.0.1:3000
tls {
issuer acme {
email user@example.com
dns googleclouddns {
gcp_project my-gcp-project
}
}
}
}
Can I use local or remote Terraform Modules in Coder templates?
One way is to reference a Terraform module from a GitHub repo to avoid duplication and then just extend it or pass template-specific parameters/resources:
# template1/main.tf
module "central-coder-module" {
source = "github.com/org/central-coder-module"
myparam = "custom-for-template1"
}
resource "ebs_volume" "custom_template1_only_resource" {
}
# template2/main.tf
module "central-coder-module" {
source = "github.com/org/central-coder-module"
myparam = "custom-for-template2"
myparam2 = "bar"
}
resource "aws_instance" "custom_template2_only_resource" {
}
Another way using local modules is to symlink the module directory inside the
template directory and then tar the template.
ln -s modules template_1/modules
tar -cvh -C ./template_1 | coder templates <push|create> -d - <name>
References:
Can I run Coder in an air-gapped or offline mode? (no Internet)?
Yes, Coder can be deployed in air-gapped or offline mode.
Our product bundles with the Terraform binary so assume access to terraform.io during installation. The docs outline rebuilding the Coder container with Terraform built-in as well as any required Terraform providers.
Direct networking from local SSH to a Coder workspace needs a STUN server. Coder defaults to Google's STUN servers, so you can either create your STUN server in your network or disable and force all traffic through the control plane's DERP proxy.
Create a randomized computer_name for an Azure VM
Azure VMs have a 15 character limit for the computer_name which can lead to
duplicate name errors.
This code produces a hashed value that will be difficult to replicate.
locals {
concatenated_string = "${data.coder_workspace.me.name}+${data.coder_workspace_owner.me.name}"
hashed_string = md5(local.concatenated_string)
truncated_hash = substr(local.hashed_string, 0, 16)
}
Do you have example JetBrains Gateway templates?
In August 2023, JetBrains certified the Coder plugin signifying enhanced stability and reliability.
The Coder plugin will appear in the Gateway UI when opened.
Selecting the most suitable template depends on how the deployment manages
JetBrains IDE versions. If downloading from
jetbrains.com is
acceptable, see the example templates below which specifies the product code,
IDE version and build number in the
coder_app
resource.
This will present an icon in the workspace dashboard which, when selected, will look for a locally installed Gateway, and open it.
Alternatively, the IDE can be baked into the container image and manually open Gateway (or
IntelliJ which has Gateway built-in), using a session token to Coder and then
open the IDE.
What options do I have for adding VS Code extensions into code-server, VS Code Desktop or Microsoft's Code Server?
Coder has an open-source project called
code-marketplace which is a
private VS Code extension marketplace. There is even integration with JFrog
Artifactory.
You can also use Microsoft's code-server - which is like Coder's, but it can connect to Microsoft's extension marketplace so Copilot and chat can be retrieved there.
Another option is to use VS Code Desktop (local) and that connects to Microsoft's marketplace.
I want to run Docker for my workspaces but not install Docker Desktop
Colima is a Docker Desktop alternative.
This example is meant for a users who want to try out Coder on a macOS device.
Install Colima and docker with:
brew install colima
brew install docker
Start Colima:
colima start
Start Colima with specific compute options:
colima start --cpu 4 --memory 8
Starting Colima on a M3 MacBook Pro:
colima start --arch x86_64 --cpu 4 --memory 8 --disk 10
Colima will show the path to the docker socket so we have a community template that prompts the Coder admin to enter the Docker socket as a Terraform variable.
How to make a coder_app optional?
An example use case is the user should decide if they want a browser-based IDE like code-server when creating the workspace.
-
Add a
coder_parameterwith typeboolto ask the user if they want the code-server IDEdata "coder_parameter" "code_server" { name = "Do you want code-server in your workspace?" description = "Use VS Code in a browser." type = "bool" default = false mutable = true icon = "/icon/code.svg" order = 6 } -
Add conditional logic to the
startup_scriptto install and start code-server depending on the value of the addedcoder_parameter# install and start code-server, VS Code in a browser if [ ${data.coder_parameter.code_server.value} = true ]; then echo "🧑🏼💻 Downloading and installing the latest code-server IDE..." curl -fsSL https://code-server.dev/install.sh | sh code-server --auth none --port 13337 >/dev/null 2>&1 & fi -
Add a Terraform meta-argument
countin thecoder_appresource so it will only create the resource if thecoder_parameteristrue# code-server resource "coder_app" "code-server" { count = data.coder_parameter.code_server.value ? 1 : 0 agent_id = coder_agent.coder.id slug = "code-server" display_name = "code-server" icon = "/icon/code.svg" url = "http://localhost:13337?folder=/home/coder" subdomain = false share = "owner" healthcheck { url = "http://localhost:13337/healthz" interval = 3 threshold = 10 } }
Why am I getting this "remote host doesn't meet VS Code Server's prerequisites" error when opening up VSCode remote in a Linux environment?
It is because, more than likely, the supported OS of either the container image or VM/VPS doesn't have the proper C libraries to run the VS Code Server. For instance, Alpine is not supported at all. If so, you need to find a container image or supported OS for the VS Code Server. For more information on OS prerequisites for Linux, please look at the VSCode docs. https://code.visualstudio.com/docs/remote/linux#_local-linux-prerequisites
How can I resolve disconnects when connected to Coder via JetBrains Gateway?
If your JetBrains IDE is disconnected for a long period of time due to a network change (for example turning off a VPN), you may find that the IDE will not reconnect once the network is re-established (for example turning a VPN back on). When this happens a persistent message will appear similar to the below:
No internet connection. Changes in the document might be lost. Trying to reconnect…
To resolve this, add this entry to your SSH config file on your local machine:
Host coder-jetbrains--*
ServerAliveInterval 5
This will make SSH check that it can contact the server every five seconds.
If it fails to do so ServerAliveCountMax times (3 by default for a total of 15 seconds) then it will close the connection which forces JetBrains to recreate the hung session.
You can tweak ServerAliveInterval and ServerAliveCountMax to increase or decrease the total timeout.
Note that the JetBrains Gateway configuration blocks for each host in your SSH config file will be overwritten by the JetBrains Gateway client when it re-authenticates to your Coder deployment so you must add the above config as a separate block and not add it to any existing ones.
How can I restrict inbound/outbound file transfers from Coder workspaces?
In certain environments, it is essential to keep confidential files within
workspaces and prevent users from uploading or downloading resources using tools
like scp or rsync.
To achieve this, template admins can use the environment variable
CODER_AGENT_BLOCK_FILE_TRANSFER to enable additional SSH command controls.
This variable allows the system to check if the executed application is on the
block list, which includes scp, rsync, ftp, and nc.
resource "docker_container" "workspace" {
...
env = [
"CODER_AGENT_TOKEN=${coder_agent.main.token}",
"CODER_AGENT_BLOCK_FILE_TRANSFER=true",
...
]
}
Important Notice
This control operates at the ssh-exec level or during sftp sessions. While
it can help prevent automated file transfers using the specified tools, users
can still SSH into the workspace and manually initiate file transfers. The
primary purpose of this feature is to warn and discourage users from downloading
confidential resources to their local machines.
Agent API capabilities, such as file operations performed by AI agents in Coder Agents chat (including chat file uploads into the workspace), are not affected by this setting because they do not go through the SSH transfer path.
For more advanced security needs, consider adopting an endpoint security solution.
How do I change the access URL for my control plane?
You may want to change the default domain that's used to access coder, i.e. yourcompany.coder.com and find yourself unfamiliar with the process.
To change the access URL associated with your control plane, you can edit any of the following variables:
- CLI using the
--access-urlflag - YAML using the
accessURLoption - or ENV using the
CODER_ACCESS_URLenvironmental variable.
For example, if you're using an environment file to configure your control plane, you'll want to edit the file located at /etc/coder.d/coder.env and edit the following:
CODER_ACCESS_URL=https://yourcompany.coder.com to your new desired URL.
Then save your changes, and reload daemon-ctl using the following command:
systemctl daemon-reload
and restart the service using:
systemctl restart coder
After coder restarts, your changes should be applied and should reflect in the admin settings.

