# MSK Maintenance (In Progress)

We provide our cBioPortal's maintenance page publicly in the hope that it may be useful to others

# Database Migration

In the database migration process, we are going to have two main steps: building importers and updating database scheme.

# Building Importers

Codebases:
Building single importer:
- Take the cbioportal hash and add it to genome nexus annotation pipeline POM -> push to fork
- Take the cbioportal hash and new genome nexus annotation pipeline hash (from above) and propagate to pipelines POM; also update db version if db migration is part of change
- Login to pipelines server
- Sets general env variables needed for building
  - Run sh /data/portal-cron/scripts/automation-environment.sh
- Specific to cbioportal build, looks for properties and xml files off of $PORTAL_HOME
  - Run export PORTAL_HOME=/data/portal-cron/git-repos/cbioportal
- Make sure the local cbioportal codebase is on the correct git hash at /data/portal-cron/git-repos/cbioportal
- Specify importer to be build
  - Copy importer properties to /data/portal-cron/git-repos/cbioportal/src/main/resources (e.g. triage /data/portal-cron/git-repos/pipelines-configuration/properties/import-triage/*)
- Navigate to pipelines folder
  - cd /data/portal-cron/git-repos/pipelines
- Build importer
  - mvn clean install -DskipTests
Build multiple importers
- Take the cbioportal hash and add it to genome nexus annotation pipeline POM -> push to fork
- Take the cbioportal hash and new genome nexus annotation pipeline hash (from above) and propagate to pipelines POM; also update db version if db migration is part of change
- Login to pipelines server
- Sets general env variables needed for building
  - Run sh /data/portal-cron/scripts/automation-environment.sh
- Specific to cbioportal build, looks for properties and xml files off of $PORTAL_HOME
  - Run export PORTAL_HOME=/data/portal-cron/git-repos/cbioportal
- Make sure the local cbioportal codebase is on the correct git hash at /data/portal-cron/git-repos/cbioportal
- Go to build importer jars folder
  - cd /data/portal-cron/git-repo/pipelines-configuration/build-importer-jars
- (Optional) Remove existing jars
  - rm *.jar
- Build all importers at once (build all importers except cmo-pipelines)
  - sh buildproductionjars.sh -sd=true -sgp=true -b=importers
  - Available parameters:
    - [--cbioportal-git-hash|-cgh=<cbioportal_commit_hash>]
    - [--skip-deployment|-sd=<true|false>]
    - [--skip-git-pull|-sgp=<true|false>]
    - [--build|-b=<build_specifier>], build_specifier should be one of the following
      - all (build for all artifacts)
      - importers (all importers except cmo-pipelines)
      - cmo-pipelines (cmo-pipelines artifacts only)
      - triage-cmo-importer
      - msk-dmp-importer
      - msk-cmo-importer
      - public-importer
      - genie-aws-importer
      - genie-archive-importer
      - hgnc-importer

# Updating Database Scheme

Database needs to be updated one by one, we have four main databases: triage, private, genie, and public. Take triage database as an example.

Migrate one database (e.g. triage)
- SSH into pipeline server
- Checkout to the commit that contains the latest database scheme
- Check if property sets up correctly to the right database (triage)
  - vi /data/portal-cron/git-repos/cbioportal/src/main/resources/application.properties
- Move to directory
  - cd /data/portal-cron/git-repos/cbioportal
- Run database migration using script:
  - python3 core/src/main/scripts/migrate_db.py --properties-file src/main/resources/application.properties --sql db-scripts/src/main/resources/migration.sql
- Monitor the DB migration process and look for possible errors
- Access database and verify the DB scheme is updated