Skip to content

Latest commit

 

History

History
181 lines (135 loc) · 11.6 KB

local-testing.md

File metadata and controls

181 lines (135 loc) · 11.6 KB
Raw

Local testing

Prerequisites

To run an Orion instance locally, first make sure you have docker and docker-compose installed on your machine.

In this tutorial we'll be using a pre-built joystream/orion docker image available on Docker Hub.

We'll start with the example configuration files provided inside the examples/ directory, ie.:

To begin, copy those files to a chosen location on your machine.

Initial configuration

You'll need to set some of the initial configuration values inside the .env file. The file already includes detailed comments explaining the meaning of each environment variable.

First, change the "placeholder values" (ie. the values inside angle brackets) such as <YOUR GATEWAY NAME>, to the actual values you wish to use.

An example configuration may look like this:

# ==================================================================================
# ========================= INITIAL GATEWAY CONFIGURATION ==========================
# ==================================================================================
APP_NAME=MyGateway
SUPPORT_NO_CATEGORY_VIDEOS=false
SUPPORT_NEW_CATEGORIES=false
KILL_SWITCH_ON=true
VIDEO_VIEW_PER_USER_TIME_LIMIT=300
# Random value generated using `openssl rand -base64 18`
OPERATOR_SECRET=zEdvAF3hhVmxo1LF3cRQTr4+
RELEVANCE_WEIGHTS="[1, 0.03, 0.3, 0.5]"
VIDEO_RELEVANCE_VIEWS_TICK=10
MAX_CACHED_ENTITIES=1000
# Random value generated using `openssl rand -base64 18`
APP_PRIVATE_KEY=QWFbJcw7FHYC9eL40ctF8sns
SESSION_EXPIRY_AFTER_INACTIVITY_MINUTES=60
SESSION_MAX_DURATION_HOURS=720
EMAIL_CONFIRMATION_ROUTE=NOT_USED
EMAIL_CONFIRMATION_TOKEN_EXPIRY_TIME_HOURS=24
EMAIL_CONFIRMATION_TOKEN_RATE_LIMIT=5
ACCOUNT_OWNERSHIP_PROOF_EXPIRY_TIME_SECONDS=300
# Random value generated using `openssl rand -base64 18`
COOKIE_SECRET=upxE4P4KZtmCTkzUnKVha741

Starting the services

Once you're done with the initial configuration, you can start the local Orion instance with:

docker-compose up -d

By default (in the example config provided), Orion will use Joystream mainnet blockchain archive gateway hosted by Subsquid (ie. https://joystream.archive.subsquid.io/graphql) for the purpose of fetching the on-chain event data. If you wish, you can set a different Subsquid archive endpoint by changing the value of the ARCHIVE_GATEWAY_URL variable in the .env file.

It will usually take 5-10 minutes for the Orion event processor to process all of the events from the Joystream blockchain and for the Orion database state to be in sync with the current blockchain state.

To track the progress of the event processor, you can use the following command:

docker-compose logs -f --tail 10 orion-processor

Once the processor is fully synced, you should start seeing log messages like those:

orion-processor      | {"level":2,"time":1685005084439,"ns":"sqd:processor","msg":"2393740 / 2393740, rate: 8806 blocks/sec, mapping: 1097 blocks/sec, 2427 items/sec, ingest: 481 blocks/sec, eta: 0s"}
orion-processor      | {"level":2,"time":1685005089928,"ns":"sqd:processor:mapping","msg":"Saving database updates..."}
orion-processor      | {"level":2,"time":1685005089934,"ns":"sqd:processor","msg":"2393741 / 2393741, rate: 8593 blocks/sec, mapping: 949 blocks/sec, 2277 items/sec, ingest: 370 blocks/sec, eta: 0s"}
orion-processor      | {"level":2,"time":1685005095428,"ns":"sqd:processor:mapping","msg":"Saving database updates..."}
orion-processor      | {"level":2,"time":1685005095433,"ns":"sqd:processor","msg":"2393742 / 2393742, rate: 8392 blocks/sec, mapping: 466 blocks/sec, 1763 items/sec, ingest: 118 blocks/sec, eta: 0s"}
orion-processor      | {"level":2,"time":1685005100938,"ns":"sqd:processor:mapping","msg":"Saving database updates..."}
orion-processor      | {"level":2,"time":1685005100943,"ns":"sqd:processor","msg":"2393743 / 2393743, rate: 8174 blocks/sec, mapping: 45 blocks/sec, 0 items/sec, ingest: 6 blocks/sec, eta: 0s"}

You can then start interacting with the APIs.

Interacting with Orion APIs

Orion consists of 2 main APIs:

  • GraphQL API - this is the main API used for executing GraphQL queries and mutations. For example: fetching data about channels, videos, data objects etc. or performing operator actions, such as exluding content, setting supported categories or modyfing the gateway configuration.
  • Auth API - this is the API responsible for the user authentication. It is also used for creating new user accounts and changing the most important gateway account settings, like the blockchain account associated with given gateway account. Full description of the Auth API, including many technical details, can be found in the Authentication API documentatin inside the developer guide.

Authentication

The first step before you will be able to interact with the GraphQL API is to authenticate with the Auth API. You can either authenticate as a Gateway Operator (knowing the OPERATOR_SECRET value from the .env file) or as an anonymous user without any special privileges.

If the OPENAPI_PLAYGROUND variable in the .env file is set to true, you can use the http://localhost:4074/playground/ endpoint to interact with the Auth API and perform the authentication.

Let's select the POST /anonymous-auth endpoint and click "Try it out":

image

In order to authenticate as operator, replace "userId": "string" with "userId": "<YOUR OPERATOR SECRET>", where <YOUR OPERATOR SECRET> should be the value of the OPERATOR_SECRET variable from the .env file, and click "Execute":

image

In order to authenticate as a standard, unprivileged user, you can just remove the "userId": "string" line completely and then click "Execute" again.

GraphQL API

In order to interact with the GraphQL API, go to http://localhost:4350/graphql. If you already authenticated with the Auth API, you will see a list of available queries on the left side of the screen.

You can click on one of them and it will be expanded, showing the list of available arguments that can be provided to this query and the list of fields that can be retrieved through it. You can simply click on an argument in order to provide a value for it and select fields you wish to be included in the response and the playground will automatically generate the GraphQL query for you. You can also modify/write the query manually if you wish. Once your query is ready, click on the "▶" button and you will see the query result on the right side of the screen:

image

As a gateway operator, you will also have access to multiple GraphQL mutations that you can use to perform various administrative tasks, such as excluding or featuring content, setting supported categories or modyfing the gateway configuration.

In order to see the list of all available mutations, first select the "Mutation" option under "Add new" on the bottom-left side of the screen and click the "+" button:

image

The first mutation you will usually want to execute when setting up a new gateway is the setSupportedCategories mutation. To read more about setting supported categories, see Setting supported categories tutorial.

Below you can find a list of all tutorials that explain how to perform various administrative tasks using the GraphQL API:

Email notifications scheduler

In order to allow notification delivery via email, follow the steps detailed in Email Notifications.

Testing with Atlas

You can test your local Orion instance with the local Atlas dev server, launched via yarn atlas:dev.

Here's a step-by-step guide on how you can configure your local environment to achieve this:

  1. First make sure the following env variables are set in your .env file:

    # Dev settings
# WARNING: DO NOT USE THESE SETTINGS IN PRODUCTION!
ORION_ENV=development
DEV_DISABLE_SAME_SITE=true

    As indicated in the comment, these settings should only be used for testing purposes and should never be used in production. They turn off some security features related to cookies that would make it more difficult to test Orion w/ Atlas locally when enabled, but which are very important in production to ensure the security of your Gateway users.

  2. Clone the Atlas repository if you haven't done so already:

    git clone https://github.com/Joystream/atlas.git
cd atlas

  3. Install Atlas dependencies (you may need to install Yarn first)

    # In case you don't have Yarn installed you can install it with volta:
# curl https://get.volta.sh | bash
# source ~/.bash_profile || source ~/.profile || source ~/.bashrc || :
# volta install yarn

yarn

  4. Adjust Atlas environment settings. To do that you can edit .env file inside packages/atlas/src directory of the Atlas repositry. You'll need to change the following values:

    1. Set VITE_ENV value to local
    2. Under Local development env URLs set VITE_LOCAL_ORION_AUTH_URL to your local Orion Auth API endpoint (http://localhost:4074/api/v1 by default)
    3. Set VITE_LOCAL_ORION_URL to your local Orion GraphQL API endpoint (http://localhost:4350/graphql by default)
    4. Set VITE_LOCAL_QUERY_NODE_SUBSCRIPTION_URL to your local Orion GraphQL API WebSocket endpoint (ws://localhost:4350/graphql by default)
    5. Unless you're running Joystream node locally, you can set VITE_LOCAL_NODE_URL to a public Joystream mainnet node endpoint (wss://rpc.joystream.org:9944) and VITE_LOCAL_FAUCET_URL to a public Joystream mainnet faucet endpoint (https://faucet.joystream.org/member-faucet/register).

  5. Start the Atlas dev server:

    yarn atlas:dev

    You should see a message similar to this one:

    image

    If you now open http://localhost:3000/ in your browser you should see the homepage of the Atlas app connected to your local Orion instance. If instead you see that the page never loads, please refer to https://vitejs.dev/guide/troubleshooting.html#requests-are-stalled-forever for troubleshooting tips.

You can now test how the changes you make to your local Orion instance configuration affect Atlas! For example, you can change supported categories, feature or exclude content or adjust video relevance weights and other configuration values and then see how these changes affect the content displayed in Atlas before you decide to make them in production.