Search Indexer Migration
The files inside the migration folder are build into multiple entry files an run in series inside separate initContainers in the cluster. The migration files have several tasks:
  • Keep the dictionaries inside AWS up to date with the repo
  • Manage files inside S3 buckets
  • Create packages inside AWS ES
  • Associate and disassociate packages with AWS ES domains
  • Keep ES indexes up to date
  • Create new indices when dictionary or index templates are updated
  • Migrate data from datasources into current index
  • Ensure builds with a faulty index don't spin up
  • Ensure content status dashboard is up to date
You can run the migration script locally to manage your dev-service instance of elasticsearch. The migrate script assumes you have dictionary files inside elasticsearch config and hence is unlikely to work with standard instances of elasticsearch

Quick start


You can run the migration with
yarn nx run services-search-indexer:migrate
This migrates the ES indexes to the latest version defined by the content-search-index-manager library. It also imports all kibana saved objects that are in ./config/kibana folder.

Sync Kibana

You can run a kibana sync with
yarn nx run services-search-indexer:migrate --sync-kibana
A sync fetches saved objects from a local running kibana instance and updates your local kibana files. It uses the ids of the objects inside the ./config/kibana folder and searches for them on your instance and updates them accordingly. Remember to run migrate before running sync-kibana.