Skip to content

Latest commit

 

History

History
63 lines (43 loc) · 3.72 KB

README.md

File metadata and controls

63 lines (43 loc) · 3.72 KB
Raw

New!

We're using Sequelize ORM

id fields are by default of type Sequelize.DataTypes.BIGINT

By default allowNull is false, put allowNull: true specifically if you want a field to be Nullable

Make migrations from diff of the models: ./point makemigration --datadir ~/.point/test1

Migrate: npm run db:migrate

If you're operating on a json field, such as storage_link.segments_sent, methods such as .push will not work, the model only understands that there has been changes if you trigger the direct assignment, i.e. storage_link.segments_sent = new_segments_sent, which will trigger the sequelize getter; Moreover, new_segments_sent cannot just be the same array with pushed items, it should be shallow copied (such as through spread operator [...original]) in order for sequelize to understand it's a different object;

Use model.refresh() to reload latest changes from the database if some other thread could have written into it

Use transactions with locking for update to avoid race conditions.

Postgres

To test out Sequelize and Postgres using Docker:

  1. Optional - rebuild the docker image image so that it uses the latest packages: docker build -t "pointnetwork/pointnetwork_node:dev" .
  2. Run pointnetwork docker services in dev mode: docker-compose -f docker-compose.yaml -f docker-compose.dev.yaml up -d
  3. Follow the logs of the website_owner container: docker-compose logs -f website_owner
  4. Follow the logs of the storage_provider container: docker-compose logs -f storage_provider
  5. Connect one terminal to website_owner node: docker exec -it pointnetwork_website_owner bash
  6. Run the database migrations for the website_owner postgres db using ./point migrate
  7. Connect another terminal to storage_provider node: docker exec -it pointnetwork_storage_provider bash
  8. Run the database migrations for the storage_provider postgres db using ./point migrate
  9. Now within the running container terminal for website_owner node run a simple example deployment: ./scripts/deploy-sites.sh hello.point --contracts

Check data in Postgres

  1. Start a new psql terminal: make psql_console
  2. Within the running psql terminal:
    1. Check the schemas are migrated, for example the files table: \d files
    2. Check the data has been saved in point_website_owner db, for example load some rows from the files table: select id, size, redundancy, expires, autorenew from files;
  3. Open the Point Network PgAdmin Dashboard and use the credentials as defined in the docker-compose.yml file.
  4. Navigate to the Point Network Database Server, select the database, select the table and from the context menu select 'View/Edit Data -> All Rows'.

Check data in LevelDB

  1. Check all the files in the website_owner (or other) LevelDB instance:
    1. Stop the website_owner node: docker-compose stop website_owner. This is because LevelDB only accepts one concurrent connection.
    2. Start the website_owner node overriding the entrypoint: docker-compose -f docker-compose.yaml -f docker-compose.dev.yaml run --entrypoint bash website_owner
    3. Run the LevelDB playground script: node ./scripts/db/playground.js
    4. Adjust the script to output the data that you are insterested in.
    5. Exit the container.
    6. Restart and reattach to the website_owner service (follow the steps at the start of this document).

Clear all data from Postgres and LevelDB to run a clean test

In order to clear all the data from both Postgres and LevelDB, you simply need to drop all docker volumes, like so:

docker-compose down
docker system prune --volumes -f

Now you are ready to test a deployment with an empty LevelDB and Postgres database.