Design a Distributed Cache (Memcache)/Playground
A look-aside cache at Facebook's scale
Click a part to change it, put something in front of it, or kill it. Turn the traffic up. Every number moves as you go; nothing is graded. Billions of reads a second fleet-wide; scaled down here to one cluster, with a 1% miss rate assumed.
Running
- Requests failing
- 0%
- Backlog growing
- none
- Instances running
- 877
What goes through it
| Operation | Offered | Outcome | Wait | Up |
|---|---|---|---|---|
| Key reads | 1.5M/s | All served | 214 ms | 99.999% |
| Misses to MySQL | 15k/s | All served | 6 ms | 99.999% |
| Writes | 3,000/s | All served | 13 ms | 99.999% |
| Invalidationsbackground | 3,000/s | Keeping up | – | – |
Wait is the expected time a caller waits; Up is the share of time every part it waits on is running.
What your changes did
Nothing yet. This is the reference design: change something to see what it buys and costs.
At this traffic
Everything is running within capacity.
Ask AI what your design does
AIThe numbers above come from a simple model. The AI reads your design, what you changed and these numbers, and explains where it holds and where it fails, citing how the engineers who built it did it. It can be wrong.