Performance

This guide covers performance optimization strategies for Teleportal applications.

Rate Limiting

Rate limiting prevents abuse and ensures fair resource usage:

const server = new Server({
  rateLimitConfig: {
    maxMessages: 100,
    windowMs: 1000,
    trackBy: "user",
  },
});

See Rate Limiting Guide for details.

VirtualStorage for Write Buffering

VirtualStorage adds batching and buffering to any DocumentStorage implementation, improving write performance:

import { VirtualStorage } from "teleportal/storage";

// Wrap any existing storage with batching
const batchedStorage = new VirtualStorage(existingDocumentStorage, {
  batchMaxSize: 100, // Batch every 100 updates
  batchWaitMs: 2000, // Or every 2 seconds
});

// Use as normal DocumentStorage
await batchedStorage.handleUpdate("doc1", update);
const doc = await batchedStorage.getDocument("doc1"); // Flushes pending writes

Benefits:

• Faster writes: Updates are buffered and batched
• Reduced I/O: Fewer database calls
• Read consistency: Pending writes are flushed on reads

When to use:

• High-frequency collaborative updates
• Slow storage backends (remote DBs, object storage)
• Write-heavy applications

Storage Optimization

Use Appropriate Storage Backend

Choose storage based on your needs:

• Redis: Fast, in-memory, good for high-frequency updates
• PostgreSQL: Reliable, transactional, good for complex queries
• S3: Scalable, cost-effective, good for large files

Indexed Keys vs Key Scanning

For unstorage-based storage:

const { documentStorage } = createUnstorage(storage, {
  scanKeys: false, // Use indexed keys (better for Redis, Memcached)
  // or
  scanKeys: true, // Use key scanning (better for PostgreSQL, MySQL)
});

Connection Optimization

Connection Reuse

Multiple providers can share the same connection:

const connection = new WebSocketConnection({ url: "wss://..." });

const provider1 = new Provider({ client: connection, document: "doc-1" });
const provider2 = new Provider({ client: connection, document: "doc-2" });

Document Switching

Use switchDocument() to efficiently switch documents:

// Reuses connection instead of creating a new one
const newProvider = provider.switchDocument({ document: "new-doc" });

Message Batching

Batch multiple messages together to reduce network overhead:

import { getBatchingTransform } from "teleportal/transports";

const batchedTransport = getBatchingTransform({
  maxBatchSize: 10,
  maxBatchDelay: 100,
});

Monitoring

Monitor performance using built-in metrics:

const metrics = await server.getMetrics();
const status = await server.getStatus();

Key metrics to watch:

• teleportal_message_duration_seconds: Message processing time
• teleportal_storage_operation_duration_seconds: Storage operation time
• teleportal_sessions_active: Active session count
• teleportal_clients_active: Active client count

Next Steps

• Scaling - Learn about scaling strategies
• Monitoring - Set up monitoring