Preventing cache stampede with Redis locks and probabilistic early expiration

The cache stampede problem

A cache stampede (dog-pile effect) happens when a popular cache key expires and many concurrent requests all miss and simultaneously query the database.

Mutex lock approach

async function getProductSafe(id) {
  const key = `product:${id}`;
  const cached = await client.get(key);
  if (cached) return JSON.parse(cached);

  const lockKey = `lock:${key}`;
  const acquired = await client.set(lockKey, '1', { NX: true, EX: 5 });

  if (!acquired) {
    // Another process is fetching — wait and retry
    await new Promise(r => setTimeout(r, 50));
    return getProductSafe(id);
  }

  const data = await db.getProduct(id);
  await client.setEx(key, 300, JSON.stringify(data));
  await client.del(lockKey);
  return data;
}

Probabilistic early expiration (XFetch)

Recompute the cache before it expires, probabilistically. The closer to expiry, the higher the chance of early refresh. No locks needed.

function shouldRefresh(ttl, beta = 1) {
  const delta = Math.random() * beta * Math.log(Date.now());
  return delta > ttl;
}
// If shouldRefresh returns true, recompute before the key expires

XFetch is simpler under very high concurrency. Use mutex locks for jobs where you absolutely cannot afford duplicate DB queries.