This repo deals with the high-level configuration of Cloud Services. main.yml
contains the source of truth for CS apps, and the akamai
folder deals with updating our Akamai configuration.
These branches sync:
- ci-beta -> nightly-beta -> qa-beta -> stage-beta
- ci-stable -> nightly-stable
- qa-stable -> stage-stable
To enable a new app in our environments, you need to create configuration for it in main.yml
, and then create a PR to merge it into the ci-beta
branch. The configuration for the non-prod beta branches is kept in sync, so changes to ci-beta
will automatically be merged into nightly-beta
and qa-beta
(as mentioned above).
When you need this config added to another environment (prod-beta
, ci-stable
, qa-stable
, prod-stable
), please open another PR for that environment. If you have any concerns about this process, feel free to ping #forum-clouddot-ui on Slack for assistance.
Here is some example configuration that demonstrates the structure, using all required and optional properties:
{app_id}:
title: App Title
api:
versions:
- v1
- v2
subItems:
oneApi:
title: Some title
versions:
- v1
channel: '#some-slack-channel'
description: App Title is a cool app that does business things for its users.
deployment_repo: https://github.com/app-deployment-repo-url
disabled_on_prod: true
docs: https://link.to.docs.com/docs
permissions:
method: isOrgAdmin
apps:
- app_id_1
frontend:
title: App Title Override
paths:
- /example/path
- /another/example/path
reload: reload/path
sub_apps:
- id: app_id_1
title: Some Sub App
default: true
permissions:
method: isEntitled
args:
- insights
- id: app_id_2
title: Another Sub App
suppress_id: true
source_repo: https://github.com/app-development-repo-url
mailing_list: [email protected]
top_level: false
Each of the following properties is required for all apps:
This is your app's ID. It's used as the path to your app, and must be unique.
The main title for your app. This is what you want everyone to see when they use your app.
This is the location of your app's deployment repo (not development repo). These repos generally have build
or deploy
as a suffix.
The following properties are used if your app has an API:
This is the list of API versions your app can use. Since v1
is the default, you'll usually want at least that one defined.
If your API consists of multiple APIs, you can list them here. Each has the same signature as {app_id.api}
.
If your API is accessible on a URL other than /api/{app_id}/{versions[0]}/openapi.json
, you can change it by passing the correct name. The URL will look like /api/{apiName}/{versions[0]}/openapi.json
The following properties are used if your app has a frontend:
If you want the name of your app to appear differently on the frontend, set this property to override it.
If you want this app to use the same codebase as another existing app, set this value to the ID of that app.
To indicate chrome how to load the application for federated modules you need to pass this property. It can either be a magic link containing yourApp#./RootApp
for most applications. If you want to be more specific you can pass in module object containing appName
, scope
and module
To indicate chrome loader from what app to load your fed-mods config.
To indicate federated modules scope of your application (you can have multiple scopes per one app). This is usually your application's name (same as appName
).
To indicate which module should be loaded when rendering your app (you can have multiple modules per one scope). This is usually ./RootApp
If you have a first-level application, this field indicates which group should be managed by this module.
If your application shares same manifest as other app or your manifest is located on completely different path, you can pass the path to it in this option.
This is the list of URL paths where your app will be located.
If your app is a parent to any other apps, those apps should be listed here. Also, if your app has a parent app, or is listed under one of the top-level bundles (e.g. Insights, RHEL, Hybrid), you should add your app to the appropriate sub_apps list.
Here are some notes on defining the items in sub_apps:
- If you specify the title field, the sub-app will be self-contained, and it will not look up the app's ID elsewhere in the config.
- If the title is specified, the sub-app's path will be determined by its ID.
- If the title is not specified, the ID will be used to find the app's details in the rest of the config.
If your app will be located under some other app, but isn't managed by that app, you can use this property to override the automatic generation of the URL. This property is commonly used for Settings apps, and tells Chrome's navigation the actual URL of your app.
This property is used if the app isn't a real app on disk, and only exists for navigation purposes. This removes the app ID on the frontend so that the nav bar works as expected.
The following properties aren't required for all apps, but may still apply to your app:
The Red Hat product ID for your application. This is tied to fields on Portal Case Management for pre-filling information.
Some applications are mounted in two locations, but use the same base repo (ex. RBAC and MUA), in this case MUA needs to point to RBAC's app.info.json, so this is the base app for that url.
This is the ID of the slack channel on ansible.slack.com that you want automatic notifications to be posted to.
This is a description of your app's purpose or functionalities, which is used by some other apps.
Setting this value to true
will disable the app from deploying to Prod (and appearing in Prod). This applies to both stable
and beta
releases.
This is the link to your app's documentation.
This is the URL of the development (not deployment) repo for your app, i.e. the one you commit to.
This is the mailing list associated with your project. Used to automate email notifications.
If this is set to true
, your app will be a top-level app, which is usually reserved for bundles (Insights, RHEL, Hybrid, Openshift, etc).
Use this if your app does not have a parent app or bundle.
If you want to hide any navigational element based on some chrome's logic, this is the right property. This defines the function to be used in order to hide nav item. (Chrome's list of methods)[https://github.com/RedHatInsights/insights-chrome#permissions].
If the permissions.method
requires some arguments in order to properly work, this is how to pass them to it: an array of items.
If you want to control visibility for multiple navigation items you can specify one permission per entry and list which apps from frontend.sub_apps
should be checked.
Before you can run the property-updating script locally, you need to have access to the Akamai API.
To do this, follow the steps located here. In step 5 of this doc the guide instructs you to set the Access Level of the Diagnostics Tools API to READ_WRITE; do this but also set the Access Level of the Property Manager API (PAPI) to READ-Write. Otherwise you will not have authorization to the configurations of Cloud Services. Make sure that the .edgerc
file you create is located in your home
directory and has the credentials defined in the [default]
section of the file.
If you're able to run the sample call at the end of the doc, you should be able to run the script. If you run into issues, there may be something wrong with your .edgerc
file.
For more information on the Akamai API, read the property manager docs.
This repository has a webhook that automatically builds a Jenkins job on every push. To configure this webhook, check the project's webhook settings
Testing local changes is straightforward. First, add a line like this to your insights-proxy spandx config:
--- a/profiles/local-frontend.js
+++ b/profiles/local-frontend.js
@@ -9,5 +9,6 @@ routes[`/beta/${SECTION}/${APP_ID}`] = { host: `http://localhost:${FRONTEND_PORT
routes[`/${SECTION}/${APP_ID}`] = { host: `http://localhost:${FRONTEND_PORT}` };
routes[`/beta/apps/${APP_ID}`] = { host: `http://localhost:${FRONTEND_PORT}` };
routes[`/apps/${APP_ID}`] = { host: `http://localhost:${FRONTEND_PORT}` };
+routes[`/beta/config`] = { host: `http://localhost:8889` };
module.exports = { routes };
Restart your insights-proxy to pick up the change.
Create a beta/config
directory inside of cloud-services-config
and copy main.yml
to it. Then, from the cloud-services-config
dir, run npx http-server -p 8889
. In your browser, go to https://ci.foo.redhat.com:1337/beta/rhel/dashboard
. You should see something logged like this from npx:
$ npx http-server -p 8889
npx: installed 25 in 2.484s
Starting up http-server, serving ./
Available on:
http://127.0.0.1:8889
http://192.168.0.25:8889
http://10.10.122.158:8889
Hit CTRL-C to stop the server
[Tue Nov 05 2019 09:50:55 GMT-0500 (Eastern Standard Time)] "GET /beta/config/main.yml" "Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:69.0) Gecko/20100
101 Firefox/69.0"
Before you go developing, make sure you can make a simple change and see it in the web UI. Try renaming "Dashboard" to "XDashboardX". To do this, make an edit to main.yml
similar to this (make sure you are editing the one in beta/config
):
diff --git a/main.yml b/main.yml
index 090fd7e..a680d06 100644
--- a/main.yml
+++ b/main.yml
@@ -152,7 +152,7 @@ cost-management:
mailing_list: [email protected]
dashboard:
- title: Dashboard
+ title: XDashboardX
channel: '#flip-mode-squad'
deployment_repo: https://github.com/RedHatInsights/insights-dashboard-build
frontend:
Then, reload the site. You may not see your change at this point! Try clearing your local storage in your browser. To do this in Firefox, hit Shift-F9 and click "Local Storage", then right click on https://ci.foo.redhat.com:1337 and delete all. Refresh the page and you should then see your changes. You'll notice too that SimpleHTTPServer logged another request. You will need to repeat this cache clearing step whenever you make changes to main.yml
in your local environment.