Preparing the distribution: VM and cluster requirements
Tip
Before you install DataLens On-premises, make sure to prepare the environment properly. This directly impacts the system’s health and performance.
Conditions for getting access to the distribution
The DataLens On-premises distribution is a commercial product. To get it, you need to accept the agreement with Yandex Cloud. Upon acceptance, you will get a unique download link. To access Yandex Cloud and download the distribution, you need an active billing account.
Tip
If you do not have a Yandex Cloud account yet, sign up for one here.
Minimum test configuration requirements
The minimum configuration that is required to run all services should be enough for testing, pilot projects, or to get your first experience with the product.
Supported operating systems
- Ubuntu 20.04+
- Debian 10+
- CentOS 7+
- Astra Linux 1.7
- Red Hat Enterprise Linux
- RED OS
Minimum test environment requirements
- vCPU: 8 vCPUs
- RAM: 16 GB
- SSD: 60 GB
- Add the
--cpu-scale 0.5flag
This configuration is suitable for multiple users but will be less performant under a load. However, it is enough to try out DataLens Enterprise.
Required resources for production use
For production environments, the requirements are higher and depend on the number of simultaneously active users and dashboard complexity.
Recommended system requirements
- vCPU: 16 vCPUs
- RAM: 32 GB
- SSD: 80 GB
Scaling
- Users
- The more users access dashboards at the same time, the higher the load on the Data API (request generation and proxying) and Control API (metadata handling). If you are planning resources to support 100+ active users, start with at least 16 vCPUs and 32 GB RAM. To get exact numbers, perform load testing.
- PostgreSQL service database
- For production environments, we recommend using an external fault-tolerant PostgreSQL cluster (master + replica). This is critical to make sure your entire system is reliable. The minimum requirements for such a cluster are 4 vCPUs, 8 GB RAM, and 100 GB SSD.
Summary
You have reviewed the requirements for you future environment and learned how to download the distribution. You also now understand the difference between testing and production configurations and how important it is to use an external PostgreSQL cluster in production.