How would you implement a simple in-process file-based cache in Node.js, and what are the tradeoffs versus using Redis?

A file-based cache writes computed results to disk (keyed by a hash of the input, often as the filename) and checks for an existing cached file before recomputing -- simple to implement with no external dependency, and durable across process restarts unlike an in-memory cache, but slower than Redis for high-frequency access due to filesystem I/O overhead, and doesn't naturally support features like automatic TTL expiration or being shared efficiently across multiple server instances the way Redis does.

const crypto = require('node:crypto');

async function getCached(key, computeFn) {
  const cacheFile = path.join('.cache', crypto.createHash('md5').update(key).digest('hex'));
  try {
    return JSON.parse(await fs.promises.readFile(cacheFile, 'utf-8'));
  } catch {
    const result = await computeFn();
    await fs.promises.writeFile(cacheFile, JSON.stringify(result));
    return result;
  }
}