Starting a grid
Whether you need to start one grid node or many, the process of starting a grid is the same.
The last step of your test plan launcher gives you the ability to launch a new grid if you don't already have one running for your account.
You can also manage your grids from your menu bar
Configuring a grid
We ask you a few simple questions, like how many nodes and which region you want to start the grid in.
We also give you some lifecycle options for the grid, when it should start and when it should stop.
Hit Launch Grid and we'll automatically spin up the infrastructure for you. When the nodes are ready they'll change from orange to green, but you can select your recently launched grid anyway.
Grids take around 5 minutes to start and are automatically stopped based on your settings or you an stop them earlier if you happen to finish your testing for the day (or night!)
A bit more information about grids
Behind every Flood test is a grid of distributed nodes, which execute your test plan and produce the results in real time.
These nodes as we refer to them, can be launched in multiple geographic regions around the world.
A grid is simply what we call a collection of nodes. You can launch up to 90 nodes per grid, and you can have as many grids as you like running your test.
Each grid node will run the exact same copy of the test plan and whatever configuration you set for the Flood.
So for example, if your load test is configured with 1,000 users and you run that on a 10 node grid, then you will effectively be generating 10,000 users from that grid. More if you start allocating more nodes or more grids to your test.
How many grid nodes do I need?
Everyone's requirements are different, and everyone can expect different levels of concurrency and throughput from a single grid node, depending on their test plan complexity and of course how well their target site performs under load.
For that reason we advise customers first run their load test on a single grid node in a geographic region that is close to their target site. We advise a planning figure of 1,000 threads per node but this can be higher or lower depending on the test plan tool, complexity and target site performance.
We call that process baselining a single node. Once you have an idea of a single node's performance you can then plan to scale out with as many nodes in as many grids or regions as you require.
A single node's performance is representative of a node in a grid of up to 90 nodes. We've designed grid nodes to be autonomous, loosely coupled and shared nothing which affords you the greatest unhindered scale when planning your load tests.