couchbase-magma

Use Couchbase Magma for large datasets needing lower memory overhead and better scale, while couchstore fits smaller datasets with fewer memory constraints. Base your storage engine selection on dataset size, write rate, and available memory.

Select 1024 vBuckets for larger Magma datasets to ensure proper distribution, while 128 vBuckets suits smaller deployments. This bucket configuration setting is immutable and must be chosen correctly during initial bucket creation.

Magma compaction behavior impacts operational performance and disk space reclamation, requiring specific tuning for your write rate. Understanding compaction settings helps manage storage overhead and maintain consistent read and write performance.

Magma requires less memory than couchstore for large datasets, but you must calculate exact memory requirements based on dataset size and write rate before bucket creation. Proper memory sizing ensures optimal storage backend performance.

Migration from couchstore to Magma faces constraints because storage backend and vBucket count are immutable bucket creation settings. You must plan a migration scenario to transfer data to a new Magma-backed bucket rather than switching engines in place.