Orakl Network
English
Search
K

Helm Chart Infra Setup

Description

This section describes how to install Orakl Network Services through the Helm Chart. This document uses helm chart to deploy in cloud (AWS, GCP) environment.
  • Kubernetes
  • Postgresql (V.14)
  • Redis (latest)
  • Persistent Volumes

Installation

01. Register the Orakl Network helm chart repository

helm repo add orakl https://helm.orakl.network
helm repo update

02. Create namespace for kubernetes

kubectl create namespace orakl
kubectl create namespace redis

03. Storage

AWS Installation
  • You must first create an efs file storage in AWS and have an ID.
helm install storage -n orakl orakl/orakl-log-aws-storage \
--set "config.efsFileSystemId=${your efs file system id}" \
--set "config.region=${your aws resion}" \
--set "config.size=100Gi"
GCP Installation
  • In GCP, you must first create a file storage and have a name for the storage
helm install storage -n orakl orakl/orakl-log-gcp-storage \
--set "config.size=100Gi" \
--set "config.pdName=${your gcp storage name}"

04. Api

Before installing the API, you must have a postgresql database ready.
helm install api -n orakl orakl/orakl-api \
--set "global.config.DATABASE_URL=${your database url}" \
--set "global.config.ENCRYPT_PASSWORD=${your key password}" \
--set "global.config.APP_PORT=3030"
  • DATABASE_URL : postgresql://{user_id}:{user_password}@{host_name}:{port}/{database_name}
  • Typically, we'll create a database called orakl and proceed from there
  • make sure for security reason not to use admin account. create new user for the database.
  • ENCRYPT_PASSWORD : You can set the key to any key you want as the conversion key. When the reporter's private key is stored in the database, it is encrypted and stored with this key.

05. Fetcher

To install Orakl Fetcher, you need to have Redis set up beforehand.
helm install fetcher -n orakl orakl/orakl-fetcher \
--set "global.config.APP_PORT=4040" \
--set "global.config.REDIS_HOST=${your redis host address }" \
--set "global.config.REDIS_PORT=${your redis port | 6379}" \
--set "global.config.ORAKL_NETWORK_API_URL=http://orakl-api.orakl.svc.cluster.local:3030/api/v1"
  • ORAKL_NETWORK_API_URL: This is the address to communicate with the Pod you installed in step 03. It uses the usual Kubernetes domain for communication, which is http://{api pod's service name}.{namespace}.svc.cluster.local:3030/api/v1 So if you installed in a namespace other than the orakl namespace, the address of the API server will need to be changed, for example, if the namespace is default, it will be http://orakl-api.default.svc.cluster.local:3030/api/v1
  • REDIS_HOST, REDIS_PORT: You can use a fresh deployment of Redis in Kubernetes, or you can use your own Redis if you have one

06. Cli

helm install cli -n orakl orakl/orakl-cli \
--set "global.config.ORAKL_NETWORK_API_URL=http://orakl-api.orakl.svc.cluster.local:3030/api/v1" \
--set "global.config.ORAKL_NETWORK_FETCHER_URL=http://orakl-fetcher.orakl.svc.cluster.local:4040/api/v1" \
--set "global.config.ORAKL_NETWORK_DELEGATOR_URL=http://orakl-delegator.orakl.svc.cluster.local:5050/api/v1"
  • ORAKL_NETWORK_API_URL, ORAKL_NETWORK_FETCHER_URL : These 2 URLs are the same convention we used to deploy the fetcher earlier, just make sure to include the service name and namespace.
  • ORAKL_NETWORK_DELEGATOR_URL : This URL will be available soon. The parts related to Delegator are in the process of being migrated.

07. Basic Configuration

If you look at the aggregator Baobab at https://config.orakl.network/, there are 12 contracts and each contract has an address. There will be more later. We have to have 12 accounts to whitelist in this contract if you want to activate reporters account. To activate your account, kindly reach out to us at [email protected].
  • list all pods in orakl namespace
    kubectl get pods -n orakl
  • access to cli pod
    kubectl exec -it cli-7b8df47f-bdlzv -n orakl -- /bin/bash
    You can set the default settings through the orakl cli part. Here we'll show you an example.
  • network setting
    yarn cli chain insert --name baobab
  • network list a
    yarn cli chain list
  • The result should be
    root@cli-7b8df47f-bdlzv:/app# yarn cli chain list
    yarn run v1.22.19
    $ node --no-warnings --experimental-specifier-resolution=node --experimental-json-modules dist/index.js chain list
    [ { id: '1', name: 'baobab' } ]
    Done in 0.70s.
  • setting service
    yarn cli service insert --name DATA_FEED
  • service check
    yarn cli service list
  • the result should be
    root@cli-7b8df47f-bdlzv:/app# yarn cli service list
    yarn run v1.22.19
    $ node --no-warnings --experimental-specifier-resolution=node --experimental-json-modules dist/index.js service list
    [ { id: '1', name: 'DATA_FEED' } ]
    Done in 0.66s.
  • setting listener
    To set up a listener, you need to know the contract address of the listener associated with the DATA FEED. In this article, we will focus on BNB-USDT. This contract is already deployed on Bisonai and the contract address can be found in the Address section of the Aggregator Baobab in the Orakl Config documentation.
    yarn cli listener insert \
    --chain baobab \
    --service DATA_FEED \
    --address 0x731a5afb6e021579138ea469b25c2ab46ff44199 \
    --eventName NewRound
  • setting adapter
    yarn cli adapter insert \
    --source https://config.orakl.network/adapter/bnb-usdt.adapter.json
    --chain baobab
    source address can be found Orakl Config
  • setting aggregator
    yarn cli aggregator insert \
    --source https://config.orakl.network/aggregator/baobab/bnb-usdt.aggregator.json \
    --chain baobab
    source address can be found Orakl Config
  • setting reporter
    yarn cli reporter insert \
    --chain baobab \
    --service DATA_FEED \
    --address ${your reporter account address} \
    --privateKey ${Your reporter account private key}
    --oracleAddress 0x731a5afb6e021579138ea469b25c2ab46ff44199
    oracleAddress refer BNB-USDT contract address
  • setting delegator
    yarn cli delegator organizationInsert --name ${your organazation name}
    yarn cli delegator reporterInsert \
    --address ${your reporter account address of BNB-USDT} \
    --organizationId 1
    The reporter address should be match with your BNB-USDT reporter's account address. The reporter's account address must be lowercase. Even if the reporter's account address is already mixed with capitalization, please make it lowercase here. We'll fix this in the future (issue caused by switching from ethers.js to caver.js)

    yarn cli delegator contractInsert \
    --address "0x731a5afb6e021579138ea469b25c2ab46ff44199"
    The contract address is BNB-USDT's contract

    yarn cli delegator functionInsert \
    --name "submit(uint256,int256)" \
    --contractId 1
    contractId should match with contract id which deployed. If you want to check contract ID inserted in database yarn cli delegator contractList. You can find out id from list.

    yarn cli delegator contractConnect \
    --contractId 1 \
    --reporterId 1
    This is the task that connects the contract and the reporter together. This can be used in the future if the contract and reporter are different or change. Reporter's id can be found yarn cli delegator reporterList

    Enable Fetcher with aggregatorHash
    yarn cli fetcher start --id ${aggregatorHash} --chain ${chainName}

08. Datafeed (aggregator)

helm install aggregator -n orakl orakl/orakl-aggregator \
--set "global.config.CHAIN=${your network name}" \
--set "global.config.LOG_LEVEL=info" \
--set "global.config.HEALTH_CHECK_PORT=8080" \
--set "global.config.NODE_ENV=production" \
--set "global.config.ORAKL_NETWORK_API_URL=http://orakl-api.orakl.svc.cluster.local:3030/api/v1" \
--set "global.config.ORAKL_NETWORK_DELEGATOR_URL=http://orakl-delegator.orakl.svc.cluster.local:5050/api/v1" \
--set "global.config.PROVIDER_URL=${your network provider url}" \
--set "global.config.REDIS_HOST=${your redis host}" \
--set "global.config.REDIS_PORT=${your redis port}" \
--set "global.config.SLACK_WEBHOOK_URL=${your slack hook url}"
  • CHAIN: baobab | cypress
  • ORAKL_NETWORK_API_URL, ORAKL_NETWORK_DELEGATOR_URL : These 2 URLs are the same convention we used to deploy the fetcher earlier, just make sure to include the service name and namespace.
  • PROVIDER_URL : The JSON RPC address of the node to use ex) https://api.baobab.klaytn.net:8651
  • SLACK_WEBHOOK_URL : If there is no value, you can leave it blank.
  • REDIS_HOST, REDIS_PORT: You can use a fresh deployment of Redis in Kubernetes, or you can use your own Redis if you have one