Performance and Caching

Strapi is fast out of the box for small datasets. At scale -- thousands of entries, deep relations, high traffic -- you need explicit caching, query optimization, and infrastructure tuning.

Performance bottleneck overview

The three most common bottlenecks:

1. Deep population -- every populate adds joins
2. N+1 queries -- fetching relations one-by-one instead of batched
3. No response caching -- identical requests hit the database every time

---

Response caching

In-memory cache (simple, single-instance)

Good for small projects, no external dependencies:

// src/api/article/middlewares/cache.js
const cache = new Map();

module.exports = (config, { strapi }) => {
  const ttl = config.ttl || 60000;        // Default 60 seconds
  const maxSize = config.maxSize || 1000;  // Max cached entries

  return async (ctx, next) => {
    // Only cache GET requests
    if (ctx.method !== 'GET') {
      await next();
      // Invalidate cache on writes
      cache.clear();
      return;
    }

    const key = `${ctx.url}|${ctx.get('Authorization') || 'anon'}`;
    const cached = cache.get(key);

    if (cached && Date.now() - cached.timestamp < ttl) {
      ctx.body = cached.body;
      ctx.status = cached.status;
      ctx.set('X-Cache', 'HIT');
      return;
    }

    await next();

    if (ctx.status === 200 && ctx.body) {
      // Evict oldest entry if cache is full
      if (cache.size >= maxSize) {
        const oldest = cache.keys().next().value;
        cache.delete(oldest);
      }

      cache.set(key, {
        body: ctx.body,
        status: ctx.status,
        timestamp: Date.now(),
      });
      ctx.set('X-Cache', 'MISS');
    }
  };
};

Redis cache (production, multi-instance)

npm install ioredis

// src/api/article/middlewares/redis-cache.js
const Redis = require('ioredis');

let redis;

module.exports = (config, { strapi }) => {
  const ttl = config.ttl || 60;  // seconds
  const prefix = config.prefix || 'strapi:cache:';

  if (!redis) {
    redis = new Redis(process.env.REDIS_URL || 'redis://localhost:6379');
  }

  return async (ctx, next) => {
    if (ctx.method !== 'GET') {
      await next();
      // Invalidate related cache on writes
      const pattern = `${prefix}${ctx.url.split('?')[0]}*`;
      const keys = await redis.keys(pattern);
      if (keys.length) await redis.del(...keys);
      return;
    }

    const key = `${prefix}${ctx.url}`;

    try {
      const cached = await redis.get(key);
      if (cached) {
        ctx.body = JSON.parse(cached);
        ctx.set('X-Cache', 'HIT');
        ctx.set('X-Cache-TTL', ttl);
        return;
      }
    } catch (err) {
      strapi.log.warn('[redis-cache] Read error:', err.message);
    }

    await next();

    if (ctx.status === 200 && ctx.body) {
      try {
        await redis.setex(key, ttl, JSON.stringify(ctx.body));
      } catch (err) {
        strapi.log.warn('[redis-cache] Write error:', err.message);
      }
      ctx.set('X-Cache', 'MISS');
    }
  };
};

Applying the cache middleware

// src/api/article/routes/article.js
const { createCoreRouter } = require('@strapi/strapi').factories;

module.exports = createCoreRouter('api::article.article', {
  config: {
    find: {
      middlewares: [
        { name: 'api::article.redis-cache', config: { ttl: 300 } },  // 5 min
      ],
    },
    findOne: {
      middlewares: [
        { name: 'api::article.redis-cache', config: { ttl: 600 } },  // 10 min
      ],
    },
  },
});

Cache invalidation via Document Service middleware

// src/index.js
const Redis = require('ioredis');
const redis = new Redis(process.env.REDIS_URL);

module.exports = {
  register({ strapi }) {
    strapi.documents.use(async (context, next) => {
      const result = await next();

      if (['create', 'update', 'delete', 'publish', 'unpublish'].includes(context.action)) {
        const prefix = 'strapi:cache:';
        const apiName = context.uid.split('.').pop();
        const pattern = `${prefix}/api/${apiName}*`;

        const keys = await redis.keys(pattern);
        if (keys.length) {
          await redis.del(...keys);
          strapi.log.debug(`[cache] Invalidated ${keys.length} keys for ${context.uid}`);
        }
      }

      return result;
    });
  },
};

---

Database optimization

Adding database indexes

Strapi doesn't add indexes automatically (beyond primary keys). For frequently filtered fields, add them manually:

// database/migrations/2025.01.15T00.00.00.add-indexes.js
module.exports = {
  async up(knex) {
    // Index on slug for lookups
    await knex.schema.alterTable('articles', (table) => {
      table.index('slug');
    });

    // Index on published_at for publication queries
    await knex.schema.alterTable('articles', (table) => {
      table.index('published_at');
    });

    // Composite index for locale + slug queries
    await knex.schema.alterTable('articles', (table) => {
      table.index(['locale', 'slug']);
    });

    // Index on foreign keys (Strapi doesn't always add these)
    await knex.schema.alterTable('articles_author_lnk', (table) => {
      table.index('article_id');
      table.index('author_id');
    });
  },

  async down(knex) {
    await knex.schema.alterTable('articles', (table) => {
      table.dropIndex('slug');
      table.dropIndex('published_at');
      table.dropIndex(['locale', 'slug']);
    });
  },
};

Connection pooling

// config/env/production/database.js
module.exports = ({ env }) => ({
  connection: {
    client: 'postgres',
    connection: {
      host: env('DATABASE_HOST'),
      port: env.int('DATABASE_PORT', 5432),
      database: env('DATABASE_NAME'),
      user: env('DATABASE_USERNAME'),
      password: env('DATABASE_PASSWORD'),
    },
    pool: {
      min: env.int('DATABASE_POOL_MIN', 2),
      max: env.int('DATABASE_POOL_MAX', 10),
      acquireTimeoutMillis: 30000,
      idleTimeoutMillis: 30000,
    },
  },
});

---

Query optimization

Select only needed fields

// BAD: fetches all 30 fields for each article
const articles = await strapi.documents('api::article.article').findMany({
  populate: '*',
});

// GOOD: only fetch what the frontend needs
const articles = await strapi.documents('api::article.article').findMany({
  fields: ['title', 'slug', 'publishedAt'],
  populate: {
    author: { fields: ['name'] },
    cover: { fields: ['url', 'alternativeText', 'width', 'height'] },
  },
});

Avoid deep population

// BAD: 4 levels of nesting
const articles = await strapi.documents('api::article.article').findMany({
  populate: {
    author: {
      populate: {
        company: {
          populate: {
            employees: {
              populate: ['avatar'],
            },
          },
        },
      },
    },
  },
});

// GOOD: fetch only what's needed, make separate queries if needed
const articles = await strapi.documents('api::article.article').findMany({
  fields: ['title', 'slug'],
  populate: {
    author: { fields: ['name', 'avatar'] },
  },
});

Pagination: always paginate

// BAD: loads all articles into memory
const all = await strapi.documents('api::article.article').findMany({});

// GOOD: paginate
const page1 = await strapi.documents('api::article.article').findMany({
  page: 1,
  pageSize: 25,
});

---

CDN integration

Serving media through a CDN

// config/middlewares.js
module.exports = [
  // ...
  {
    name: 'strapi::security',
    config: {
      contentSecurityPolicy: {
        directives: {
          'img-src': ["'self'", 'data:', 'cdn.example.com', '*.cloudinary.com'],
        },
      },
    },
  },
  // ...
];

Setting cache headers for API responses

// src/middlewares/cache-headers.js
module.exports = (config, { strapi }) => {
  const maxAge = config.maxAge || 60;

  return async (ctx, next) => {
    await next();

    if (ctx.method === 'GET' && ctx.status === 200) {
      // Public endpoints can be cached by CDN
      if (!ctx.state.user) {
        ctx.set('Cache-Control', `public, max-age=${maxAge}, s-maxage=${maxAge * 2}`);
        ctx.set('Vary', 'Accept, Accept-Encoding');
      } else {
        ctx.set('Cache-Control', 'private, no-cache');
      }
    }
  };
};

Stale-while-revalidate pattern

ctx.set(
  'Cache-Control',
  `public, max-age=60, s-maxage=300, stale-while-revalidate=600`
);
// CDN serves stale content for up to 10 minutes while revalidating in the background

---

Monitoring and profiling

Request timing middleware

// src/middlewares/request-timing.js
module.exports = (config, { strapi }) => {
  const slowThreshold = config.slowThreshold || 1000;

  return async (ctx, next) => {
    const start = Date.now();
    await next();
    const duration = Date.now() - start;

    ctx.set('X-Response-Time', `${duration}ms`);

    if (duration > slowThreshold) {
      strapi.log.warn({
        msg: 'Slow request',
        method: ctx.method,
        url: ctx.url,
        status: ctx.status,
        duration: `${duration}ms`,
        query: ctx.query,
      });
    }
  };
};

Database query logging

// config/database.js
module.exports = ({ env }) => ({
  connection: {
    client: 'postgres',
    connection: { /* ... */ },
    debug: env.bool('DATABASE_DEBUG', false),  // Logs all SQL queries
  },
});

Enable temporarily with DATABASE_DEBUG=true to find slow queries.

---

Performance checklist

Action	Impact	Effort
Add fields to every query	High -- reduces payload 50-80%	Low
Add database indexes on filtered columns	High -- 10x faster lookups	Low
Implement response caching (Redis)	Very high -- eliminates DB hits	Medium
Set Cache-Control headers	High -- enables CDN caching	Low
Paginate all list endpoints	High -- prevents memory spikes	Low
Limit population depth	Medium -- fewer joins	Low
Use connection pooling	Medium -- prevents connection exhaustion	Low
Profile slow requests	Varies -- identifies the real bottleneck	Medium

---

Common pitfalls

Pitfall	Problem	Fix
populate=* in production	Fetches entire relational graph	Explicitly list fields and relations
No cache invalidation	Stale data served forever	Invalidate on write operations
Cache key ignores auth	User A sees user B's data	Include auth state in cache key
SQLite in production	Locks under concurrent writes	Use PostgreSQL or MySQL
No database indexes	Full table scans on every query	Add indexes on slug, publishedAt, foreign keys
Logging every query in production	I/O overhead from debug logging	Only enable debug temporarily

---

See also

• Relations and Population -- population strategies that directly impact performance
• Configuration and Deployment -- database and server configuration
• GraphQL Customization -- GraphQL-specific performance patterns (DataLoader)
• Middleware and Policies -- applying cache middleware to routes