replit-performance-tuning
Optimize Replit app performance: cold start, memory, Nix caching, and deployment speed. Use when experiencing slow startup, high memory usage, deployment timeouts, or optimizing Replit container resource usage. Trigger with phrases like "replit performance", "optimize replit", "replit slow", "replit cold start", "replit memory", "replit startup time".
Allowed Tools
Provided by Plugin
replit-pack
Claude Code skill pack for Replit (30 skills)
Installation
This skill is included in the replit-pack plugin:
/plugin install replit-pack@claude-code-plugins-plus
Click to copy
Instructions
Replit Performance Tuning
Overview
Optimize Replit app performance across the entire lifecycle: cold start reduction, Nix environment caching, build speed, runtime memory management, and deployment configuration. Replit containers have resource limits — efficient usage is critical.
Prerequisites
- Replit app deployed or running in Workspace
- Understanding of
.replitandreplit.nix - Access to deployment monitoring
Instructions
Step 1: Reduce Cold Start Time
Autoscale deployments scale to zero when idle. First request triggers a cold start (10-30s). Minimize it:
// 1. Lazy-load heavy modules — only import when needed
// BAD: imports everything at startup
import { heavyAnalytics } from './analytics'; // 500ms
import { imageProcessor } from './images'; // 300ms
// GOOD: import on demand
app.get('/api/analyze', async (req, res) => {
const { heavyAnalytics } = await import('./analytics');
res.json(await heavyAnalytics.process(req.query));
});
// 2. Defer non-critical initialization
let dbPool: Pool | null = null;
function getDB(): Pool {
if (!dbPool) {
dbPool = new Pool({
connectionString: process.env.DATABASE_URL,
ssl: { rejectUnauthorized: false },
max: 5, // Keep pool small for faster init
});
}
return dbPool;
}
// 3. Start server immediately, initialize after
const app = express();
const PORT = parseInt(process.env.PORT || '3000');
app.listen(PORT, '0.0.0.0', () => {
console.log(`Server ready in ${process.uptime().toFixed(1)}s`);
// Warm up in background after server is accepting requests
warmup().catch(console.error);
});
async function warmup() {
await getDB().query('SELECT 1'); // Pre-connect
}
Step 2: Optimize Nix Environment
# replit.nix — only include what you actually need
# BAD: kitchen-sink approach
{ pkgs }: {
deps = [
pkgs.nodejs-20_x
pkgs.python311
pkgs.go
pkgs.rustc
pkgs.cargo
pkgs.postgresql
pkgs.redis
pkgs.imagemagick
];
}
# GOOD: minimal deps for a Node.js app
{ pkgs }: {
deps = [
pkgs.nodejs-20_x
pkgs.nodePackages.typescript-language-server
];
# Only add postgresql if you need psql CLI:
# pkgs.postgresql
}
# .replit — pin Nix channel for cache hits
[nix]
channel = "stable-24_05"
# Changing channel invalidates all Nix caches
# Only upgrade when needed
Step 3: Optimize Build Step
# .replit — fast production builds
[deployment]
build = ["sh", "-c", "npm ci --production && npm run build"]
run = ["sh", "-c", "node dist/index.js"]
// package.json — optimize build scripts
{
"scripts": {
"build": "tsc --incremental",
"start": "node dist/index.js",
"dev": "tsx watch src/index.ts"
}
}
// tsconfig.json — incremental builds
{
"compilerOptions": {
"incremental": true,
"tsBuildInfoFile": ".tsbuildinfo",
"skipLibCheck": true
}
}
Tips for faster builds:
- Use
npm ci(notnpm install) — deterministic, faster - Add
--productionto skip devDependencies - Use TypeScript
--incrementalfor rebuild caching - Avoid
postinstallscripts that compile native addons
Step 4: Memory Management
Replit containers have memory limits (512 MB to 16 GiB depending on plan/tier):
// Monitor memory usage
function logMemory() {
const usage = process.memoryUsage();
const mb = (bytes: number) => Math.round(bytes / 1024 / 1024);
console.log({
heapUsed: `${mb(usage.heapUsed)} MB`,
heapTotal: `${mb(usage.heapTotal)} MB`,
rss: `${mb(usage.rss)} MB`,
external: `${mb(usage.external)} MB`,
});
}
// Check every 60 seconds
setInterval(logMemory, 60000);
// Expose via health endpoint
app.get('/health', (req, res) => {
const mem = process.memoryUsage();
res.json({
status: 'ok',
uptime: process.uptime(),
memoryMB: Math.round(mem.heapUsed / 1024 / 1024),
memoryPercent: ((mem.heapUsed / mem.heapTotal) * 100).toFixed(1),
});
});
Memory optimization patterns:
// Stream large files instead of loading into memory
import { createReadStream } from 'fs';
app.get('/download/:file', (req, res) => {
const stream = createReadStream(`/tmp/${req.params.file}`);
stream.pipe(res);
});
// Paginate database queries
app.get('/api/items', async (req, res) => {
const page = parseInt(req.query.page as string) || 1;
const limit = 50;
const offset = (page - 1) * limit;
const { rows } = await pool.query(
'SELECT * FROM items ORDER BY id LIMIT $1 OFFSET $2',
[limit, offset]
);
res.json({ items: rows, page, hasMore: rows.length === limit });
});
// Clear caches when memory is high
const cache = new Map<string, any>();
setInterval(() => {
if (process.memoryUsage().heapUsed > 400 * 1024 * 1024) {
cache.clear();
console.log('Cache cleared due to high memory');
}
}, 30000);
Step 5: Database Connection Efficiency
// PostgreSQL pool tuning for Replit
const pool = new Pool({
connectionString: process.env.DATABASE_URL,
ssl: { rejectUnauthorized: false },
max: 5, // Small pool — containers are limited
idleTimeoutMillis: 30000, // Close idle connections after 30s
connectionTimeoutMillis: 5000,
});
// Use connection pooling, never create per-request connections
// BAD: new Pool() per request
// GOOD: single pool, shared across requests
Step 6: Deployment Type Selection
| Scenario | Best Type | Why |
|---|---|---|
| < 100 daily requests | Autoscale | Free when idle |
| Consistent traffic | Reserved VM | No cold starts |
| Static frontend | Static | Fastest, cheapest |
| Latency-sensitive API | Reserved VM | Always warm |
| Cron jobs / webhooks | Reserved VM | Must be always-on |
Error Handling
| Issue | Cause | Solution |
|---|---|---|
| Cold start > 15s | Heavy imports | Lazy-load, defer init |
| OOM killed | Exceeding memory limit | Stream data, reduce pool size |
| Build timeout | Slow npm install | Use npm ci --production |
| Slow first query | DB cold connection | Pre-connect in warmup() |
Resources
Next Steps
For cost optimization, see replit-cost-tuning.