A quick guide on deciding when to refresh schema and when to refresh cache in Omni.
What’s the Difference?
- Schema refresh updates Omni’s understanding of your database structure (tables, columns, views).
- Cache refresh clears or rebuilds stored query results so users see fresher data.
When to Refresh Schema
Refresh schema when the structure of your database has changed:
- A table or view was added, removed, or renamed
- Columns were added, removed, or renamed
- Data types or constraints changed
- dbt models were added or modified in a way that changes output structures
Schema refresh options:
- Soft refresh: picks up new/updated objects
- Hard refresh: also removes dropped objects
Docs for reference:
- Modeling overview: Getting started with modeling | Omni Analytics Documentation
- Schema refresh options: Model management | Omni Analytics Documentation
- dbt integration: dbt | Omni Analytics Documentation
When to Refresh Cache
Refresh or expire cache when the data has changed, not the structure:
- New data has arrived (ETL/ELT updates)
- Existing values were updated or deleted
- You need dashboards or queries to show the latest state immediately
Docs for reference:
- Caching concepts: Caching and Query Performance | Omni Analytics Documentation
- Advanced model concepts (cache policies):
Developing in Omni: Advanced modeling concepts | Omni Analytics Documentation
Quick Decision Guide
| Situation | Refresh Schema? | Refresh Cache? |
|---|---|---|
| A column/table was added or removed |  Yes |
 Not needed |
| Data updated but schema stayed the same | No |
Yes |
| dbt added a new model/table | Yes |
Optional |
| A table was dropped in the source DB | Yes (hard) |
No |
| Dashboard needs newest data ASAP | No |
Yes |
Summary
- Schema refresh = structure changes
- Cache refresh = data changes
- Use both intentionally to keep Omni fast, accurate, and up to date.