The plugin registry is automatically built for every Che release (e.g. quay.io/eclipse/che-plugin-registry:7.6.0
) in addition to nightly builds available at quay.io/eclipse/che-plugin-registry:nightly
. If a custom build of the registry is desired, there is a build.sh
script at the root of this repository that can be used to easily build the image:
Usage: ./build.sh [OPTIONS]
Options:
--help
Print this message.
--tag, -t [TAG]
Docker image tag to be used for image; default: 'nightly'
--registry, -r [REGISTRY]
Docker registry to be used for image; default 'quay.io'
--organization, -o [ORGANIZATION]
Docker image organization to be used for image; default: 'eclipse'
--latest-only
Build registry to only contain 'latest' meta.yamls; default: 'false'
--offline
Build offline version of registry, with all extension artifacts
cached in the registry; disabled by default.
--rhel
Build using the rhel.Dockerfile instead of the default
Note that the Dockerfiles in this repository utilize multi-stage builds, so Docker version 17.05 or higher is required.
Using the --offline
option in build.sh
, it's possible to build an image for the plugin registry that includes all referenced extension artifacts (i.e. all .theia
and .vsix
archives). The offline version of the plugin registry is useful in network-limited scenarios, as it avoids the need to download plugin extensions from the outside internet.
You can deploy Che plugin registry on Openshift with command.
oc new-app -f openshift/che-plugin-registry.yml \
-p IMAGE="quay.io/eclipse/che-plugin-registry" \
-p IMAGE_TAG="nightly" \
-p PULL_POLICY="IfNotPresent"
You can deploy Che plugin registry on Kubernetes using helm. For example if you want to deploy it in the namespace kube-che
and you are using minikube
you can use the following command.
NAMESPACE="kube-che"
DOMAIN="$(minikube ip).nip.io"
helm upgrade --install che-plugin-registry \
--debug \
--namespace ${NAMESPACE} \
--set global.ingressDomain=${DOMAIN} \
./kubernetes/che-plugin-registry/
You can use the following command to uninstall it.
helm delete --purge che-plugin-registry
docker run -it --rm -p 8080:8080 quay.io/eclipse/che-plugin-registry:nightly
Here is an overview of all fields that can be present in plugin meta YAML files. This document represents the current v3
version.
apiVersion: # plugin meta.yaml API version -- v2; v1 supported for backwards compatability
publisher: # publisher name; must match [-a-z0-9]+
name: # plugin name; must match [-a-z0-9]+
version: # plugin version; must match [-.a-z0-9]+
type: # plugin type; e.g. "Theia plugin", "Che Editor"
displayName: # name shown in user dashboard
title: # plugin title
description: # short description of plugin's purpose
icon: # link to SVG or PNG icon
repository: # URL for plugin (e.g. Github repo)
category: # see [1]
firstPublicationDate: # optional; see [2]
latestUpdateDate: # optional; see [3]
deprecate: # optional; section for deprecating plugins in favor of others
autoMigrate: # boolean
migrateTo: # new org/plugin-id/version, e.g. redhat/vscode-apache-camel/latest
spec: # spec (used to be che-plugin.yaml)
endpoints: # optional; plugin endpoints -- see https://www.eclipse.org/che/docs/che-6/servers.html for more details
- name:
public: # if true, endpoint is exposed publicly
targetPort:
attributes:
protocol: # protocol used for communicating over endpoint, e.g. 'ws' or 'http'
secure: # use secure version of protocol above; convert 'ws' -> 'wss', 'http' -> 'https'
discoverable: # if false, no k8s service is created for this endpoint
cookiesAuthEnabled: # if true, endpoint is exposed through JWTProxy
type:
path:
containers: # optional; sidecar containers for plugin
- image:
name: # name used for sidecar container
memoryLimit: # Kubernetes/OpenShift-spec memory limit string (e.g. "512Mi"). Refer to https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/#meaning-of-memory for details.
memoryRequest: # Kubernetes/OpenShift-spec memory request string (e.g. "256Mi"). Refer to https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/#meaning-of-memory for details.
cpuLimit: # Kubernetes/OpenShift-spec CPU limit string (e.g. "500m"). Refer to https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/#meaning-of-cpu for details.
cpuRequest: # Kubernetes/OpenShift-spec CPU request string (e.g. "125m"). Refer to https://kubernetes.io/docs/concepts/configuration/manage-compute-resources-container/#meaning-of-cpu for details.
env: # list of env vars to set in sidecar
- name:
value:
command: # optional; definition of root process command inside container
- /bin/sh
args: # optional; list arguments for root process command inside container
- -c
- ./entrypoint.sh
volumes: # volumes required by plugin
- mountPath:
name:
ephemeral: # boolean; if true volume will be ephemeral, otherwise volume will be persisted
ports: # ports exposed by plugin (on the container)
- exposedPort:
commands: # development commands available to plugin container
- name:
workingDir:
command: # list of commands + arguments, e.g.:
- rm
- -rf
- /cache/.m2/repository
mountSources: # boolean
lifecycle: # container lifecycle hooks -- see https://kubernetes.io/docs/concepts/containers/container-lifecycle-hooks/
postStart: # the postStart event immediately after a Container is started -- see https://kubernetes.io/docs/tasks/configure-pod-container/attach-handler-lifecycle-event/
exec: # Executes a specific command, resources consumed by the command are counted against the Container
command: ["/bin/sh", "-c", "/bin/post-start.sh"] # Command is the command line to execute inside the container, the working directory for the command is root ('/')
# in the container's filesystem. The command is simply exec'd, it is not run inside a shell, so traditional shell
# instructions ('|', etc) won't work. To use a shell, you need to explicitly call out to that shell. Exit status
# of 0 is treated as live/healthy and non-zero is unhealthy
# -- see https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.18/#execaction-v1-core
preStop: # the preStop event immediately before the Container is terminated -- see https://kubernetes.io/docs/tasks/configure-pod-container/attach-handler-lifecycle-event/
exec: # Executes a specific command, resources consumed by the command are counted against the Container
command: ["/bin/sh","-c","/bin/pre-stop.sh"] # Command is the command line to execute inside the container, the working directory for the command is root ('/')
# in the container's filesystem. The command is simply exec'd, it is not run inside a shell, so traditional shell
# instructions ('|', etc) won't work. To use a shell, you need to explicitly call out to that shell. Exit status
# of 0 is treated as live/healthy and non-zero is unhealthy
# -- see https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.18/#execaction-v1-core
initContainers: # optional; init containers for sidecar plugin
- image:
name: # name used for sidecar container
memorylimit: # Kubernetes/OpenShift-spec memory limit string (e.g. "512Mi")
env: # list of env vars to set in sidecar
- name:
value:
command: # optional; definition of root process command inside container
- /bin/sh
args: # optional; list arguments for root process command inside container
- -c
- ./entrypoint.sh
volumes: # volumes required by plugin
- mountPath:
name:
ephemeral: # boolean; if true volume will be ephemeral, otherwise volume will be persisted
ports: # ports exposed by plugin (on the container)
- exposedPort:
commands: # development commands available to plugin container
- name:
workingDir:
command: # list of commands + arguments, e.g.:
- rm
- -rf
- /cache/.m2/repository
mountSources: # boolean
workspaceEnv: # optional; env vars for the workspace
- name:
value:
extensions: # optional; required for VS Code/Theia plugins; list of urls to plugin artifacts (.vsix/.theia files) -- examples follow
- https://github.com/Azure/vscode-kubernetes-tools/releases/download/0.1.17/vscode-kubernetes-tools-0.1.17.vsix # example
- vscode:extension/redhat.vscode-xml # example
- https://github.com/redhat-developer/omnisharp-theia-plugin/releases/download/v0.0.1/omnisharp_theia_plugin.theia # example
- relative:extension/resources/java-0.46.0-1549.vsix # example; see [4]
1 - Category must be equal to one of the following: "Editor", "Debugger", "Formatter", "Language", "Linter", "Snippet", "Theme", "Other"
2 - firstPublicationDate is not required to be present in YAML, as if not present, it will be generated during Plugin Registry dockerimage build
3 - latestUpdateDate is not required to be present in YAML, as it will be generated during Plugin Registry dockerimage build
4 - extensions starting with relative:extension
are resolved relative to the path of index.json
-- e.g. v3
. This is primarily to support an offline or airgapped instance of the plugin registry. See Offline and airgapped registry images for details.
Note that the spec
section above comes from the older che-plugin.yaml
spec. The endpoints
, containers
, and workspaceEnv
are passed back to Che server and are used to define the sidecar that is added to the workspace.
At the moment, some of these fields (that are related to plugin viewer) are validated during the Plugin Registry dockerimage build.
Example:
curl "http://localhost:8080/v3/plugins/index.json"
or
curl "http://localhost:8080/v3/plugins/"
Response:
[
{
"id": "eclipse/che-theia/latest",
"displayName": "theia-ide",
"version": "latest",
"type": "Che Editor",
"name": "che-theia",
"description": "Eclipse Theia",
"publisher": "eclipse",
"links": {
"self": "/v3/plugins/eclipse/che-theia/latest"
}
},
{
"id": "eclipse/che-theia/next",
"displayName": "theia-ide",
"version": "next",
"type": "Che Editor",
"name": "che-theia",
"description": "Eclipse Theia, get the latest release each day.",
"publisher": "eclipse",
"links": {
"self": "/v3/plugins/eclipse/che-theia/next"
}
},
{
"id": "eclipse/x-lang-ls/2019.08.20",
"displayName": "x lang support",
"version": "2019.08.20",
"type": "VS Code extension",
"name": "x-lang-ls",
"description": "Provides support for language x",
"publisher": "eclipse",
"deprecate": {
"automigrate": true,
"migrateTo": "eclipse/x-lang-ls/2019.11.05"
},
"links": {
"self": "/v3/plugins/eclipse/x-lang-ls/2019.08.20"
}
},
{
"id": "eclipse/x-lang-ls/2019.11.05",
"displayName": "x lang support",
"version": "2019.11.05",
"type": "VS Code extension",
"name": "x-lang-ls",
"description": "Provides support for language x",
"publisher": "eclipse",
"links": {
"self": "/v3/plugins/eclipse/x-lang-ls/2019.11.05"
}
}
]
Example:
curl "http://localhost:8080/v3/plugins/eclipse/che-theia/next/meta.yaml"
or
curl "http://localhost:8080/v3/plugins/eclipse/che-theia/latest/meta.yaml"
Response:
apiVersion: v2
publisher: eclipse
name: che-theia
version: next
type: Che Editor
displayName: theia-ide
title: Eclipse Theia development version.
description: Eclipse Theia, get the latest release each day.
icon: https://raw.githubusercontent.com/theia-ide/theia/master/logo/theia-logo-no-text-black.svg?sanitize=true
category: Editor
repository: https://github.com/eclipse/che-theia
firstPublicationDate: "2019-03-07"
spec:
endpoints:
- name: theia
public: true
targetPort: 3100
attributes:
protocol: http
type: ide
secure: true
cookiesAuthEnabled: true
discoverable: false
- name: theia-dev
public: true
targetPort: 3130
attributes:
protocol: http
type: ide-dev
discoverable: false
- name: theia-redirect-1
public: true
targetPort: 13131
attributes:
protocol: http
discoverable: false
- name: theia-redirect-2
public: true
targetPort: 13132
attributes:
protocol: http
discoverable: false
- name: theia-redirect-3
public: true
targetPort: 13133
attributes:
protocol: http
discoverable: false
containers:
- name: theia-ide
image: eclipse/che-theia:next
env:
- name: THEIA_PLUGINS
value: local-dir:///plugins
- name: HOSTED_PLUGIN_HOSTNAME
value: 0.0.0.0
- name: HOSTED_PLUGIN_PORT
value: "3130"
volumes:
- mountPath: /plugins
name: plugins
mountSources: true
ports:
- exposedPort: 3100
- exposedPort: 3130
- exposedPort: 13131
- exposedPort: 13132
- exposedPort: 13133
memoryLimit: 512M
latestUpdateDate: "2019-07-05"
The following CentOS CI jobs are associated with the repository:
master
- builds CentOS images on each commit to themaster
branch and pushes them to quay.io.nightly
- builds CentOS images and pushes them to quay.io on a daily basis from themaster
branch. Thenightly
version of the plugin registry is used by default by thenightly
version of the Eclipse Che, which is also built on a daily basis by theall-che-docker-images-nightly
CI job.release
- builds CentOS and corresponding RHEL images from therelease
branch. CentOS images are public and pushed to quay.io. RHEL images are also pushed to quay.io, but to the private repositories and then used by the "Hosted Che" plugin registry - https://che-plugin-registry.openshift.io/.release-preview
- builds CentOS and corresponding RHEL images from therelease-preview
branch and automatically updates "Hosted Che" staging plugin registry deployment based on the new version of images - https://che-plugin-registry.prod-preview.openshift.io/. CentOS images are public and pushed to quay.io. RHEL images are also pushed to quay.io, but to the private repositories.
Che is open sourced under the Eclipse Public License 2.0.