BeQuery
Sign in
Glossary

Read replica

A read replica is a synchronised copy of a database used for reads, so heavy queries don't load the primary.

A read replica is a copy of a production database kept in sync with the primary, used to serve read-only queries. Analytics and reporting run against the replica so that expensive aggregations never compete with live transactions on the primary. BeQuery takes this further: instead of a live replica, it clones your store's database into an isolated PostgreSQL copy, so reporting queries have zero impact on your shop and you get a faster analytical engine.

How syncing works

Related terms

Keyset pagination

Keyset pagination pages through rows using a column cursor (e.g. last id seen) instead of OFFSET, staying fast and consistent on large tables.

EAV (Entity-Attribute-Value)

EAV is a database pattern that stores attributes as rows instead of columns — flexible, but heavy to query.