Concepts6 min read
Amdahl's law: formula, examples and scaling limits
Amdahl's law explained with the formula, a speedup table and worked examples, plus Gustafson's law, the USL and what all three mean for load tests.
loadtest.app
Load testing tools, compared and explained
Reviews, comparisons and guides for anyone who has to prove a system holds up under load. Published by LoadFocus, a load testing service, and open about it: when a competitor is the better pick, we say so.
Latest
-
How to use sysbench for CPU, disk and MySQL benchmarks
How sysbench works: CPU, memory and file I/O tests, OLTP scripts for MySQL and PostgreSQL, the options that matter, how to read the output, its limits.
Guides6 min read
-
YCSB: workloads A-F, how to run it and its limits
YCSB explained: what the Yahoo! Cloud Serving Benchmark measures, core workloads A to F, load and run phases, how to read its output and where it stops.
Guides7 min read
-
pgbench: how to benchmark and load test PostgreSQL
pgbench explained: initialize with -s, run with -c, -j and -T, write custom scripts, throttle with -R and read tps, latency and failed transactions.
Guides6 min read
-
NBomber load testing: a review of the .NET framework
NBomber load testing reviewed: C# and F# scenarios, inject and constant simulations, thresholds, the HTTP plugin, cluster mode and its company licence.
Tool reviews6 min read
-
Tsung load testing: a review of the Erlang tool
Tsung load testing reviewed: XML scenarios, arrival phases, HTTP, XMPP, database and MQTT plugins, distributed clients, reports and its 2023 last release.
Tool reviews6 min read
-
Azure Load Testing: a review of Microsoft's service
Azure Load Testing reviewed: JMeter and Locust tests, quick URL tests, engines and limits, regions, private endpoints, pricing per virtual user hour, CI.
Tool reviews6 min read
-
Apdex score: what it is and how to calculate it
Apdex score explained: the formula, how to choose the T threshold, a worked example, and how JMeter, Artillery and LoadFocus compute it in a load test.
Concepts7 min read
-
Artillery load testing: a review of the Node.js tool
Artillery load testing reviewed: YAML and TypeScript scripts, arrival-rate phases, Playwright tests, ensure checks, AWS and Azure runs and pricing.
Tool reviews6 min read
-
Tail latency: what it is and how to measure it
Tail latency explained: why the slowest 1% of requests matters, how fan-out makes it worse, what causes it, and how to measure p99 honestly in a load test.
Concepts7 min read
-
gRPC load testing: tools, setup and what to watch
gRPC load testing explained: which tools speak gRPC natively, how to write a k6 or ghz test, and why failed calls still come back as HTTP 200 responses.
Guides6 min read
-
Throughput vs latency: how they interact under load
Throughput vs latency explained: what each measures, why JMeter's Latency is not response time, and why latency climbs once throughput hits a ceiling.
Concepts7 min read
-
Gatling vs Locust: JVM simulations or Python locustfiles?
Gatling vs Locust compared on licence, runtime, test code, load model, scaling out, pass/fail checks and reports, with advice on when each one fits.
Comparisons6 min read
-
JMeter vs Postman: which tool for load testing?
JMeter vs Postman for load testing: what Postman's virtual users, cloud runs and pricing do, where they stop, and when JMeter is the better choice.
Comparisons5 min read
-
k6 vs Locust: JavaScript scripts or Python users?
k6 vs Locust compared on licence, scripting language, load models, pass/fail checks, protocols and scaling out, with advice on when each tool fits.
Comparisons6 min read
-
Gatling load testing: a review of the JVM tool
Gatling load testing reviewed: simulations, open and closed injection, checks, assertions, HTML reports, build plugins, CI, limits and when k6 fits better.
Tool reviews7 min read
-
Gatling vs JMeter: tests as code or a GUI test plan?
Gatling vs JMeter compared on licence, scripting, protocols, recorders, load models, scaling out, CI and reports, with plain advice on when each one fits.
Comparisons7 min read
-
Coordinated omission: what it is and how to avoid it
Coordinated omission explained: why a load generator that waits for slow replies hides them, how to spot it and how wrk2, k6, Gatling and JMeter avoid it.
Concepts7 min read
-
How to load test a GraphQL API: a practical guide
How to load test a GraphQL API: name every operation, mix queries and mutations, catch the errors hidden behind HTTP 200 and watch caching and N+1.
Guides6 min read
-
Gatling vs k6: JVM simulations or a single Go binary?
Gatling vs k6 compared on licence, runtime, scripting languages, protocols, load models, pass/fail checks and scaling out, with advice on when each fits.
Comparisons6 min read
-
HTTP benchmarking tools: wrk, hey, oha, ab and more
wrk, wrk2, hey, oha, ab, bombardier, autocannon and h2load compared: rate control, HTTP/2 support, latency output, licence and recent releases.
Tool reviews7 min read
-
Load test pass/fail criteria: how to set them
How to set load test pass/fail criteria: start from an SLO, choose latency, error and rate limits, split hard fails from warnings and encode them per tool.
Guides6 min read
-
Little's law in load testing: size and check your tests
Little's law in load testing explained: how users, request rate and response time fit together, how to size a closed-model test, and where it breaks.
Concepts7 min read
-
Locust load testing: a review of the Python tool
Locust load testing reviewed: Python locustfiles, web UI and headless runs, distributed mode, CI, limits, other Python tools and when k6 fits better.
Tool reviews6 min read
-
Locust vs JMeter: Python scripts or JMeter test plans?
Locust vs JMeter compared on licence, scripting, protocols, distributed runs, CI and reports, with plain advice on when each tool is the better choice.
Comparisons7 min read
-
Load testing in GitHub Actions: a practical guide
How to set up load testing in GitHub Actions with k6, Artillery, Locust or JMeter: thresholds that fail the job, artifacts and what one runner can handle.
Guides6 min read
-
Open vs closed workload model: which one should you use?
Open vs closed workload model explained: arrival rate or fixed users, why closed tests hide slowdowns, and how k6, Gatling, Locust and JMeter model each.
Concepts7 min read
-
Think time in load testing: what it is and how to set it
Think time in load testing explained: why pauses change the load your users create, think time vs pacing, and how to set it in k6, JMeter and Locust.
Concepts7 min read
-
WebSocket load testing tools: six options compared
WebSocket load testing tools compared: k6, Artillery, Gatling, JMeter, Locust and Tsung, what each supports natively and when to pick another one.
Tool reviews6 min read
-
How to run a load test: a step-by-step guide
How to run a load test from start to finish: set a goal, script one user journey, smoke test it, ramp up the load and read the results that matter.
Guides6 min read
-
k6 vs Artillery: which load testing tool fits your team?
k6 vs Artillery compared on scripting, protocols, pass/fail checks, scaling out and licences, with plain advice on when each tool is the better pick.
Comparisons6 min read
-
Open source load testing tools: six options reviewed
Six open source load testing tools reviewed: JMeter, k6, Gatling, Locust, Artillery and Vegeta, with what each is good at and when to pick another.
Tool reviews7 min read
Browse by topic
Single-tool reviews, head-to-head comparisons and step-by-step guides.
- Tool reviews Load testing tools one at a time, with what each one is good and bad at.
- Comparisons Two or more load testing tools side by side, and which one fits which team.
- Guides Load and performance testing step by step, from the first script to reading the results.
- Concepts What the numbers mean: percentiles, throughput, think time, ramp-up and the rest.
LoadFocus, the publisher of this site, runs k6 and JMeter scripts from 25 cloud regions.
Try LoadFocus