Maelstrom — load & scaling testing
Maelstrom drives realistic traffic at a target and measures how it holds up. It spins up virtual users — simulated concurrent visitors — and watches response times, error rates, and throughput climb as the load increases. The result is a clear picture of where your application slows down or starts to fail.
When to use it
Section titled “When to use it”- Before a launch, sale, or campaign, to confirm you can handle the expected spike in traffic.
- After an infrastructure or code change, to check you haven’t regressed.
- To find the point at which response times degrade or errors appear — your practical capacity ceiling.
Maelstrom is about performance and capacity, not security. To probe a target for vulnerabilities, use Vortex or network scanning instead.
Key concepts
Section titled “Key concepts”Virtual users
Section titled “Virtual users”Simulated concurrent users. The virtual-user count is the main lever for how much load you apply — more virtual users means more simultaneous requests against the target.
Load profile
Section titled “Load profile”The shape of the load over time. You choose how virtual users arrive and depart — for example a steady ramp, discrete steps, a sudden spike, or a flat hold at a fixed level. Different profiles answer different questions.
Latency percentiles
Section titled “Latency percentiles”How long requests take, reported as percentiles (p50, p90, p99) rather than a single average. The p99 tells you what your slowest users experience, which an average tends to hide.
Error-rate by load
Section titled “Error-rate by load”How the proportion of failed responses changes as the virtual-user count rises. The point where errors start climbing is usually your real capacity limit.
Where to go next
Section titled “Where to go next”- Running a load test — the run flow and every option explained.
- Reading a scaling report — what the metrics mean and how to interpret them.