The Redis caching machine.
Why is the second request faster?
Loading the experiment…
Why is the second request faster?
Fit the three devices and request a value.
A cache-aside application looks in Redis first. If the key is absent, it reads the source and stores a copy with an expiry time. A hit can skip the source read. A source-only update leaves that copy stale until it expires or is invalidated. This machine lets you compare those paths using one key and an explicit clock.
Single-key deterministic cache-aside model, separate from Redis client tracking. Logical time advances only with the clock button. The application fills a missing cache entry and explicitly invalidates it after a coordinated write. Input latencies are illustrative, not measured Redis benchmarks. GET does not renew the TTL.
Produce a miss followed by a fresh cache hit.
Compare repeated requests with fresh and stale cached values.