In complex data pipelines, change is inevitable. Whether you’re introducing new columns, adjusting existing fields, or refining your schema, traditional workflows often require tearing everything down and starting over — a costly and disruptive process. Geomi’s new Backfill tool for No-Code Indexing makes this process painless. Instead of rebuilding from scratch, you can efficiently reprocess and update your datasets to ensure your indexing stays accurate, consistent, and production-ready. We’ve published a video to guide you through the interface and demonstrate how to trigger and monitor a backfill.
This blog provides more detailed context with suggestions on best practices and insights on how and when to use backfills.
Why Does Backfilling Matter?
In data engineering, backfilling refers to the process of reprocessing data - a crucial practice for builders to ensure accuracy, reliability, and completeness in their systems.
- Evolve your data model: Add new fields and tables without losing existing data.
- Minimize disruption: Backfill data in parallel without affecting live operations or user experience.
- Recompute data with updated logic: Apply updated processing rules to past transactions, ensuring consistency across old and new data.
How to Backfill on Geomi
Creating backfills for your No Code Indexers is straightforward. Just follow these steps:
- Select your processor: From the resources page, select the processor that you want to edit.
- Make changes to your config: Update the config according to the changes you want to make to the data. Index a new event, or add/delete columns, tables or database indices. You can also edit “event field” → “table column” mappings.
- Select an upgrade option: On the next page, there will be a list of upgrade options. If your edits are compatible, the backfilling option will be recommended to you.
- Define the timeframe: Select the transaction version you want to start the backfill from.
- Run and monitor: Start the backfill and watch progress in real time, with clear feedback on job status.
- Validate the results: Confirm the updated data by querying the GraphQL API.
Best Practices for Backfill
- Target only the transactions that are necessary: Instead of defaulting to a full backfill across a large range of transactions, scope your backfill to the transaction ranges that are directly impacted by your schema or config changes. For example, if you’re adding a new event, start the backfill from the version that event was first emitted on-chain.
- Document your backfills: Keep a record of when and why you ran a backfill. It helps future troubleshooting and ensures your team has full visibility.
- Communicate with stakeholders: Let downstream users (e.g. analysts, GraphQL clients) know when a backfill is in progress so they can anticipate updates in dashboards.
Get Started
Geomi’s Backfill feature simplifies a process that’s often manual and error-prone in other platforms. By pairing the video walkthrough with the strategies above, you’ll not only know how to run a backfill – you’ll also understand how and why to use it effectively. Ready to give it a try? Jump into Geomi and explore the Backfill feature with these useful links.
👉 Check out the specific Geomi developer doc for backfilling.
👾 Join our community for support, give feedback, and participate in discussions with other developers.
📚 Explore the Geomi platform to see all available infrastructure tools.