Skip to main content

Storing large objects or BLOBs in the database

Optimize provides recommendations to help identify and resolve performance issues caused by storing large objects in the database. It also suggests alternative approaches to mitigate these challenges.

The following model uses the Bytes type:

model User {
  id    Int    @id @default(autoincrement())
  name  String?
  // Storing raw image data directly in the database
  avatarBytes Bytes?
}

What is the problem?​

Storing large binary objects (such as images) in the database can lead to several challenges:

  • Excessive storage usage: Large objects occupy significant space in the database, complicating management.
  • Increased I/O load: Handling large objects adds strain to the database's input/output operations.
  • Slower query performance: Most traditional databases are not optimized for efficiently serving large binary content, resulting in performance degradation during queries or updates.

Moreover, storing large objects directly in the database can cause backups to become disproportionately large, increasing the time required for restoration processes. Serving these files through the database also creates a performance bottleneck, particularly under high traffic or frequent access scenarios.