Skip to main content

​
Why Real Time

Once archives are loaded, Real Time Delivery pushes incremental changes so your systems track live chain activity, liquidity, and prices. The same datasets described in the Data Catalog stream in near real time with SLAs documented under Access & SLA.

​
Delivery Options

ModeDescriptionBest For
Streaming (Webhooks / gRPC)BlockDB pushes batched payloads to your endpoint.Event-driven microservices, fraud detection.
Managed buckets (hourly slices)Fresh parquet/CSV files land in a dedicated bucket.Warehouses that prefer file-based ingest.
Polling the Historic APIUse cursors on endpoints like /evm/blocks to pull continuously.Lightweight setups without streaming infra.
All options include _tracing_id so you can correlate updates back to the archive baseline.

​
Implementation Steps

  1. Finalize archive backfill (see Archive Delivery) to guarantee a consistent starting point.
  2. Select datasets & cadence using the Coverage and Data Granularity docs.
  3. Provision endpoints (webhook URL, bucket path, or API polling job) with authentication controlled by your BlockDB account team.
  4. Ingest + merge: apply CDC logic or upserts into the tables created from /BlockDb.Postgres.Tables.Public. Preserve _updated_at for reconciliation.
  5. Monitor SLAs leveraging Data Freshness and configured alerts in your observability stack.

​
Verification & Lineage

  • Use Lineage endpoints to confirm the provenance of any streaming record.
  • Trigger Verification APIs (receipt roots, logs bloom) when an alert fires or before downstream settlements.
  • Because Real Time Delivery shares the same _tracing_id space as archives, any discrepancy can be traced back to its original block, log, or price computation.

​
Operational Best Practices

  • Retry policy: implement exponential backoff and idempotent processors; BlockDB payloads include sequence numbers to guard against duplication.
  • Schema drift: subscribe to Schema Governance updates to detect new columns before they reach production feeds.
  • Testing: keep a staging environment pointed at a lower-rate stream to validate transformations before promoting to prod.