Batch Loading

Optimize lazy collection loading by batching multiple requests into a single database query.

API Reference: SyncServer · PrismaAdapter

The Problem

Without batching, lazy collections cause N+1 queries:


// 100 users displayed
const users = useCollectionModels<User>('user')
 
users.forEach(user => {
  // Each user's posts loaded separately
  user.posts.map(p => p.title)
  // 100 separate requests! 😔
})

The Solution

Batch all requests into one:


import { SyncServer } from '@gluonic/server'
 
// Server implements batchFetch
const server = SyncServer({
  database,
  auth,
  batchFetch: async (orgId, keys) => {
    // keys: [
    //   { t: 'post', indexKey: 'authorId', indexVal: 'user-1' },
    //   { t: 'post', indexKey: 'authorId', indexVal: 'user-2' },
    //   ... 98 more
    // ]
    
    // Single query for all!
    const posts = await prisma.post.findMany({
      where: {
        userId: orgId,
        authorId: { in: keys.map(k => k.indexVal) }
      }
    })
    
    // Group by authorId
    const grouped = {}
    for (const post of posts) {
      const key = `post:authorId:${post.authorId}`
      grouped[key] = grouped[key] || []
      grouped[key].push({
        t: 'post',
        id: post.id,
        v: post.version,
        p: serialize(post)
      })
    }
    
    return grouped
  }
})

Result: 1 request instead of 100! 🎉

Quick Implementation

For Prisma adapter with auto-config:


import { PrismaAdapter } from '@gluonic/server-prisma'
 
const database = PrismaAdapter({
  prisma,
  models: {
    post: {
      ownership: 'userId',
      versioning: 'version',
      batchIndexKeys: ['authorId', 'categoryId']  // Enable batch loading!
    }
  }
})
 
// Batch loading is automatic ✓

Coming soon: Automatic batch loading configuration will be enabled by default in a future release.

See full implementation guide in existing docs or coming advanced section.