Skip to content

Latest commit

 

History

History
258 lines (180 loc) · 8.99 KB

README.md

File metadata and controls

258 lines (180 loc) · 8.99 KB
Raw

Setup serratus-api development environment on MacOSX or Linux (Fedora, RHEL, CentOS)

Install Ansible dependencies on Linux

pkcon install -y git
pkcon install -y python3-pip
pkcon install -y python3-virtualenv

Install Ansible dependencies on MacOSX

brew install git python gnu-tar
pip3 install virtualenv openshift

Install the latest Python and setup a new Python virtualenv

# This step might be virtualenv-3 for you. 
virtualenv ~/python

source ~/python/bin/activate
echo "source ~/python/bin/activate" | tee -a ~/.bash_profile

Install the latest Ansible

pip install setuptools_rust wheel
pip install --upgrade pip

Install dependencies on Linux

pkcon install -y maven
pkcon install -y gcc
pkcon install -y make
pkcon install -y git
pkcon install -y bison
pkcon install -y flex
pkcon install -y readline-devel
pkcon install -y zlib-devel
pkcon install -y systemd-devel
pkcon install -y libxml2-devel
pkcon install -y libxslt-devel
pkcon install -y openssl-devel
pkcon install -y perl-core
pkcon install -y libselinux-devel
pkcon install -y container-selinux
pkcon install -y java-1.8.0-openjdk
pkcon install -y java-11-openjdk

Install dependencies on MacOSX

brew install maven

Setup Ansible

Install python3 application dependencies

pip3 install psycopg2-binary openshift

Setup the directory for the project and clone the git repository into it

install -d ~/.local/src/serratus-api
git clone [email protected]:team19hackathon2021/serratus-api.git ~/.local/src/serratus-api

Setup the Ansible Galaxy roles for installing the complete project locally.

install -d ~/.ansible/roles
git clone [email protected]:computate-org/computate_postgres.git ~/.ansible/roles/computate.computate_postgres
git clone [email protected]:computate-org/computate_zookeeper.git ~/.ansible/roles/computate.computate_zookeeper
git clone [email protected]:computate-org/computate_solr.git ~/.ansible/roles/computate.computate_solr
git clone [email protected]:computate-org/computate_project.git ~/.ansible/roles/computate.computate_project

Run the Ansible Galaxy roles to install the complete project locally.

ansible-playbook ~/.ansible/roles/computate.computate_postgres/install.yml

ansible-playbook ~/.ansible/roles/computate.computate_zookeeper/install.yml

ansible-playbook ~/.ansible/roles/computate.computate_solr/install.yml

ansible-playbook ~/.ansible/roles/computate.computate_project/install.yml -e SITE_NAME=serratus-api -e ENABLE_CODE_GENERATION_SERVICE=true

About the code generation

Much of the getters, setters, initialization code, API code, and page code is generated using this open source library:

https://github.com/computate-org/computate

It can be confusing how it works, so you can disable it by removing -e ENABLE_CODE_GENERATION_SERVICE=true in the command above. If you do wish to use the code generation service, then you will need to run the touch.sh script 4 times to initialize the code records in the Solr search engine. Every piece of the code is read and stored in the search engine, and the code learns about itself as it is modified. In order to link together the generated code between 2 foreign key related classes, first all of the base classes must be read into the search engine which might take a couple of passes of running the touch.sh script, followed by one of the model classes, followed by the other model class.

So to see the code generation working, open one terminal and watch the logs:

journalctl -f --user-unit watch-serratus-api

And in the other terminal, run the touch script at least 4 times, waiting for the logs to stop each time, until there are no more errors in the logs, or in the code:

~/.local/src/serratus-api/bin/touch.sh

You can view the code in the Solr search engine here:

http://localhost:8983/solr/#/computate/query?q=*:*&q.op=OR

For questions, please reach out with work-in-progress pull requests in Github in the project here.

Configure Eclipse

Install the Maven plugin for Eclipse

  • In Eclipse, go to Help -> Eclipse Marketplace...
  • Install "Maven Integration for Eclipse"

Import the serratus-api project into Eclipse

  • In Eclipse, go to File -> Import...
  • Select Maven -> Existing Maven Projects
  • Click [ Next > ]
  • Browse to the directory: ~/.local/src/serratus-api
  • Click [ Finish ]

Setup an Eclipse Debug/Run configuration to run and debug serratus-api

  • In Eclipse, go to File -> Debug Configurations...
  • Right click on Java Application -> New Configuration
  • Name: serratus-api QuarkusApp
  • Project: serratus-api
  • Main class: io.serratus.api.enus.vertx.QuarkusApp

In the "Arguments" tab

Setup the following VM arguments to disable caching for easier web development:

-DfileResolverCachingEnabled=false -Dvertx.disableFileCaching=true

In the Environment tab

Setup the following variables to setup the Vert.x verticle.

  • CLUSTER_PORT: 13081
  • CONFIG_PATH: ~/.local/src/serratus-api/config/serratus-api.yml
  • SITE_INSTANCES: 5
  • VERTXWEB_ENVIRONMENT: dev
  • WORKER_POOL_SIZE: 2
  • ZOOKEEPER_HOST_NAME: localhost
  • ZOOKEEPER_PORT: 2181

Click [ Apply ] and [ Debug ] to debug the application.

Deploy serratus-api to OpenShift with Ansible

To deploy serratus-api to OpenShift with Ansible, you will want to follow the instructions to install Ansible on your system first above "Install Ansible dependencies on Linux".

Setup ~/.ansible/roles directory

A default place to install Ansible roles from Ansible Galaxy is in ~/.ansible/roles. Make sure this directory exists:

install -d ~/.ansible/roles

Clone the Ansible roles for deploying the applications to OpenShift

git clone [email protected]:computate-org/computate_postgres_openshift.git ~/.ansible/roles/computate.computate_postgres_openshift
git clone [email protected]:computate-org/computate_zookeeper_openshift.git ~/.ansible/roles/computate.computate_zookeeper_openshift
git clone [email protected]:computate-org/computate_solr_openshift.git ~/.ansible/roles/computate.computate_solr_openshift
git clone [email protected]:computate-org/computate_project_openshift.git ~/.ansible/roles/computate.computate_project_openshift

Create an ansible vault for your OpenShift.

You can create and edit an encrypted ansible vault with a password for the host secrets for your shared OpenShift inventory to deploy serratus-api. It will have you create a password when you save the file for the first time, like using vim to exit.

sudo install -d -o $USER ~/.local/src/serratus-api-ansible
install -d ~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault
ansible-vault create ~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault
ansible-vault edit ~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault

The contents of the vault will contain the secrets needed to override any default values you want to change in the app defaults defined here.

https://github.com/team19hackathon2021/serratus-api/blob/main/openshift/defaults.yml

Here is an example of a vault that I have used to deploy the serratus-api application. You will want to update these values to reflect your OpenShift environment, like the REDHAT_OPENSHIFT_TOKEN which you will need to obtain after logging into OpenShift. Or the REDHAT_OPENSHIFT_STORAGE_CLASS_NAME which might be different than gp2 for you. If so, try creating a persistent volume in the UI to figure out a good storage class for your environment:

SITE_NAME: serratus-api

REDHAT_OPENSHIFT_HOST: https://api.rh-us-east-1.openshift.com
REDHAT_OPENSHIFT_TOKEN: OcrtrXzKNKVj0riR2FvfqORgGfnURx98G8zRPd2MUvs
REDHAT_OPENSHIFT_NAMESPACE: rh-impact
REDHAT_OPENSHIFT_STORAGE_CLASS_NAME: gp2

POSTGRES_DB_NAME: sampledb
POSTGRES_DB_USER: computate
POSTGRES_DB_PASSWORD: qVTaaa23aIkLmw
POSTGRES_VOLUME_SIZE: 1Gi
POSTGRES_STORAGE_CLASS_NAME: gp2

ZOOKEEPER_VOLUME_SIZE: 1Gi
ZOOKEEPER_STORAGE_CLASS_NAME: gp2

SOLR_VOLUME_SIZE: 2Gi
SOLR_STORAGE_CLASS_NAME: gp2

AUTH_REALM: TEAM19
AUTH_RESOURCE: team19
AUTH_SECRET: 0518f65a-f86d-42e8-ad65-00f46920443d
AUTH_HOST_NAME: sso.computate.org
AUTH_PORT: 443
AUTH_SSL: true
AUTH_TOKEN_URI: "/auth/realms/{{ AUTH_REALM }}/protocol/openid-connect/token"

Run the Ansible automation to deploy the applications to OpenShift

ansible-playbook --vault-id @prompt -e @~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault ~/.ansible/roles/computate.computate_postgres_openshift/install.yml

ansible-playbook --vault-id @prompt -e @~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault ~/.ansible/roles/computate.computate_zookeeper_openshift/install.yml

ansible-playbook --vault-id @prompt -e @~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault ~/.ansible/roles/computate.computate_solr_openshift/install.yml

ansible-playbook --vault-id @prompt -e @~/.local/src/serratus-api-ansible/vaults/$USER-staging/vault ~/.ansible/roles/computate.computate_project_openshift/install.yml

See the serratus-api application staged here in OpenShift

https://serratus-api.computate.org/