A node hour is the usage of our grid nodes or machines in the cloud for a total of one hour spread across all nodes. This means you can split this up any way that you want - one node hour could be:
- an hour of testing on one node
- 30 minutes of testing on two nodes
- 12 minutes of testing on five nodes
and so on.
We use the concept of charging in node hours, not number of users like other services, because this makes it very cost-effective for you. It gives you a lot more flexibility because you basically rent the machine from our partner, AWS, and max out its resources to run your tests. This allows you to scale up to a large number of users very simply and cheaply.
How many users can I run on one node?
While we don't limit the number of users you run per se, there is a limit imposed by the resources of each node. A node that runs at 100% CPU/memory utilisation, for example, is likely to be the bottleneck in the load test and will probably give inaccurate results. This can happen when trying to run too many users per node.
As a general rule, here are our recommendations for the number of users you can run for each tool that we support:
- JMeter - 1000 users
- Gatling - 1000 users
- Flood Element - 50 users
- Selenium - 5 users
Click here for more information on how many nodes you may need.
The Free plan allows you to test the service for Free and includes 5 node hours (a one-time inclusion). We restrict Free plans to using one node at a time. Paid plans do not have any such restrictions apart from the number of node hours included.