Instance Management

This guide covers managing OpenScan.AI explorer instances over their lifecycle — from launch through scaling, reindexing, and decommissioning. It applies to both hosted EaaS instances (managed with the OpenScan.AI team) and self-hosted deployments.


Instance Lifecycle

Launch

  1. Deploy the stack (Setup & Installation) or request a hosted instance via the contact form.
  2. Wait for the initial index to catch up to the chain head.
  3. Run Explorer Verification, then announce the explorer to your community.

Steady State

Scaling

  • API under load — add replicas behind a load balancer; the API is stateless.
  • Database under load — increase RAM, move to faster storage, or add read replicas.
  • Indexing too slow — check chain-node capacity first; it is usually the bottleneck.

Reindexing

Occasionally a release or a data issue requires reindexing part or all of history:

  1. Take a database backup.
  2. Trigger the reindex for the affected range (or full history) via the indexer’s admin command:
Terminal window
docker compose exec indexer ./bin/reindex --from-block 0
  1. Expect reduced data freshness until the range catches up; the explorer stays online throughout.

Multi-Instance Management

Running explorers for several chains:

  • Keep each instance’s configuration, database, and containers fully isolated.
  • Name resources per chain (openscan-xdc-postgres, openscan-wan-api) to avoid operational mistakes.
  • Upgrade one instance at a time and verify before moving on.

Decommissioning

  1. Redirect or remove DNS for the explorer domain.
  2. Take a final database backup if the data has archival value.
  3. Stop the stack and remove volumes:
Terminal window
docker compose down -v
  1. For hosted instances, coordinate shutdown with the OpenScan.AI team.