How to Run a Load Test with 50k+ Concurrent Users Using JMeter and BlazeMeter

This article explains how to conduct a smooth load test with 50,000 concurrent users using JMeter and BlazeMeter, covering script creation, local testing, sandbox testing, configuring users per engine, cluster testing, and leveraging the Master/Slave feature to achieve the target load.

Step 1 – Write your script : Obtain the latest JMeter version, install required plugins, and create the test plan either with the BlazeMeter Chrome extension, JMeter's HTTP(S) Test Script Recorder, or manually.

Step 2 – Run a local test : Use a single thread and iteration to debug with the View Results Tree, verify all scenarios, then scale to 10‑20 threads for a 10‑minute run, checking response times, errors, and throughput.

Step 3 – BlazeMeter sandbox test : Configure a sandbox test (e.g., 300 users, 1 controller, 50‑minute duration), ensure firewall CIDR lists are whitelisted, upload all required files (CSV, JAR, JSON, etc.), and avoid using absolute paths.

Step 4 – Set users per engine : With one controller and one engine, determine the maximum threads an engine can handle without exceeding ~75% CPU and ~85% memory, then adjust thread counts accordingly (e.g., 500 threads, 40‑minute ramp‑up).

Step 5 – Install and test a cluster : Build a cluster of one controller and up to 14 engines, run the test for the full duration, monitor CPU/memory limits on both engines and controller, and verify hit rates and user capacity.

Step 6 – Use Master/Slave mode : Create a Master test and multiple Slave tests (e.g., 9 slaves), configure each with the appropriate engine count, and launch the Master to run ten parallel tests of 5,000 users each, achieving the overall 50k‑user load.