More API servers vs a faster disk
More servers help only when the servers are the limit. Here the document store disk is the limit: it runs at 113 percent of its 500 MB/s throughput, and API CPU is 33 percent. Two more API servers leave p99 at 5,696 ms and 10 percent errors. A 1,000 MB/s volume gives p99 25 ms and costs less.
One CPU graph · CPU 40% · disk no graph · the fault is hidden
280 req/s · CPU 70% · mean 22 ms
Case 1
Photo pages read 64 KB documents at 8,800 reads/s. API CPU is 33 percent, and the store disk is at 113 percent of 500 MB/s. The budget is 70 USD/mo. More servers cost 56 USD/mo and the faster disk 40 USD/mo.
| Measure | Goal | Add 2 API servers | Move to a 1,000 MB/s disk |
|---|---|---|---|
| p99 latency | under 100 ms | 5.7 s Fail | 25 ms Pass |
| Error rate | under 0.1% | 10.3% Fail | 0% Pass |
| Monthly cost | 56 USD/mo | 40 USD/mo |