Cluster Configuration
A cluster in Fransys is the backbone that powers your application by distributing the workload across multiple machines. By adjusting your cluster’s configuration, you can tailor performance, availability, and resource usage to match your application’s needs.
Choosing a provider
For now, the only available provider is Hetzner, a German based company.
We will in the future support as many of the most reliable providers as possible, including :
-
Amazon Web Services
-
Google Cloud
-
Microsoft Azure
-
OVH
-
Scaleway
Choosing a Region
The region you select determines the physical location of your servers. This impacts both performance and latency, especially for users located in different parts of the world.
Clusters can be deployed across various regions, and selecting the right one depends on the proximity to your user base or specific compliance needs.
For the moment, all clusters are in the Finland (Helsinki) region. More regions will be available in future updates.
Defining the Number of Nodes
A node represents a single machine (virtual or physical) in the cluster. The more nodes you have, the more capacity and redundancy your application will have.
Use the slider to adjust the number of worker nodes in your cluster. This scales the infrastructure horizontally, providing greater resilience and handling for more concurrent users or higher workloads. Increasing the number of nodes also increases the fault tolerance, meaning if one node fails, the others can handle the workload.
Starting at 2 nodes, we will also include between 1 and 3 control plane nodes. Theses nodes are the brain of your cluster. They are fully managed and you won’t have to worry about them, but they do impact the price.
We highly recommend a minimum of 3 nodes for reliability, and a minimum of 6 for bigger clusters.
After 6 nodes the scaling will be vertical, giving each node more CPUs and RAM.
We will add support for bigger clusters and auto scaling in the near future.
Selecting Node Performance
In addition to the number of nodes, the performance of each node can be tailored. Nodes come in different configurations of CPU and RAM to suit your app’s processing and memory needs.
In this example, there are four options:
- cx22: 2 CPUs, 4GB RAM
- cx32: 4 CPUs, 8GB RAM
- cx42: 8 CPUs, 16GB RAM
- cx52: 16 CPUs, 32GB RAM
Cost Preview
As you adjust your node count and performance, the total monthly cost is displayed at the bottom of the configuration panel. Fransys updates the cost dynamically to reflect changes, giving you a clear idea of how much you’ll pay based on your current setup.
For instance, selecting one cx22 node results in a cost of 14.96€ per month, witch is the smallest you can do.
Scaling the Cluster
While Fransys supports manual scaling at the moment, meaning you adjust the cluster size yourself, automatic scaling is planned for the near future. This feature will allow your cluster to automatically adjust based on current workloads, scaling up during heavy use or scaling down when it’s less busy, saving on resources.