# How to do stress testing on my nakama node or cluster?
**URL:** https://forum.heroiclabs.com/t/how-to-do-stress-testing-on-my-nakama-node-or-cluster/5787
**Category:** Runtime Framework
**Created:** [November 9, 2024, 7:04pm UTC](https://forum.heroiclabs.com/t/how-to-do-stress-testing-on-my-nakama-node-or-cluster/5787 "2024-11-09T19:04:42Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@virtouso](https://forum.heroiclabs.com/u/virtouso)
#### Post date: [November 9, 2024, 7:04pm UTC](https://forum.heroiclabs.com/t/how-to-do-stress-testing-on-my-nakama-node-or-cluster/5787/1 "2024-11-09T19:04:42Z")
For rest APIs, it’s simple. Tools like JMeter allow you to apply all types of tests to a certain hardware configuration.
how can I check how many concurrent matches can my backend handle? why it uses sockets and there are many scenarios for a game to finish, how can I handle it? is there any good tool or technique for such scenarios?
---
### Author:  [@Eatos](https://forum.heroiclabs.com/u/Eatos)
#### Post date: [November 10, 2024, 11:40am UTC](https://forum.heroiclabs.com/t/how-to-do-stress-testing-on-my-nakama-node-or-cluster/5787/2 "2024-11-10T11:40:08Z")
To stress test your setup for concurrent matches I also haven’t found any official tool, since most of the game will be different, or may use different runtime but you could try creating a scripts to profile it.
For example:
- Script can create a socket connection with user id, queue for match maker, and join match, send few ping pongs across like 5-10mins and see what happens.
- In my case for initial profiling i created a script that initialize N number of matches, matches try to broadcast match state across 5min then closes, and i observe stats, how much memory it uses, cpu, latency and so on.
- This process didnt show hot spots or what to optimize, but it showed that overall there were issue with some of data that needed to be reconfigured if i wanted to have a lot of matches on some low configurations.