Single server

Watch the database become the bottleneck

One service in front of one database. Latency climbs sharply as the database fills up.

What to watch

One server, one database. The database is the smaller of the two: 6 requests at a time at 30ms each, so it runs out near 200 a second. Drag the load past 4x and the wait builds there first, not at the API.

What it is made of

Component	Count
Client	1
Database	1
Service	1

Under load

Measured by running this design through the simulator at multiples of its own offered rate of 50 requests a second. Twenty seconds of simulated time, one fixed seed, so the same numbers come out every time.

This design holds 150 requests a second with 0% errors. At 200 it is losing 6% of them.

Offered	Goodput	p99	Errors
50	38	116ms	0%
75	71	120.9ms	0%
100	104	123.5ms	0%
125	122	134.3ms	0%
150	139	148.3ms	0%
200	189	388.5ms	6.1%
300	177	454.3ms	44%
400	200	597.1ms	50.1%

Open this example

Load Single server from the Examples menu and drag the traffic slider yourself.