Configurations here manage both postgres database column migrations and hasura configuration in a single script that can be run in one of three environments sandbox (mainnet NEAR data), testnet (testnet NEAR data), and mainnet.
Running scripts/redeploy.sh [env] can be done locally but will require a {environment_name}.env file will need to be present. Ask if you need help with the contents of this file.
Note, local migrations should only ever be done run sandbox because they lack the inconsitency checks in CICD
Migrations under the hood use diesel, a Rust library for database manipulation. You will need to install this via cargo to test migrations locally.
At this time, environment variables for cloud triggers (ultimately this should be replaced) are required for each hasura configuration and this needs to be done manually.
Currently HASURA_POSTGRES, CLOUD_FUNCTIONS_BASE_URL, and NEAR_NETWORK must be set for metadata to apply correctly.
CICD uses a docker images with tools pre-installed (Builder.dockerfile) and deployed to ghcr.io. For more information on working with ghcr.io (Github Container Repo) see docs here
CICD will attempt to run migrations and apply hasura metadta via scripts/redploy.sh.
Whenever inconsistencies between these are detected, the run will fail.