HA Cluster is not tied to specific hardware. It runs on standard server hardware. You have full freedom for future expansions.
For the best performance, we recommend the following processors: XEON or AMD server CPU. Dual socket CPU configurations will provide you with even better performance.
For the best performance of the Enterprise Virtualization OS, it should be installed on an SSD with a minimum size of 30 GB. This installation drive is configured separately from the data drive – it is not possible to install the VMs on the installation drive.
The operating system requires its own drive, and for data, you need at least one additional drive. A data drive can be a simple hard drive or a software/hardware RAID.
It is strongly recommended to use server-grade hard drives. Desktop/notebook hard drives are not recommended.
Hybrid cache functionality allows a mix of slow SATA or SAS hard drives with SSD or NVMe hard drives.
The server should have at least 16 GB of RAM.
You need at least 2 network cards. Multiple, teamed network ports provide higher redundancy and performance.
To achieve optimal performance, it is recommended to use 10 Gbit network cards, especially for mirroring. To save on a switch, it is possible to connect them directly.
The cluster distinguishes between heartbeat and replication network.
All network controllers not used for replication will be used for communication with the outside world and for communication between the nodes.
Network cards used for mirroring between the servers are part of the replication network.
For the best performance and high availability, the server should have 4 individual network interface cards with a minimum bandwidth of 1 Gigabit (10 Gigabit recommended). They should be configured as two separately connected networks. One of the connections should be connected to the corporate network. The other connection should be connected to the other server on a private replication network. For the replication network, we recommend a 10 Gigabit network.
If dual-port cards are used, each port on the card should be assigned to a separate bond so that the failure of one card does not result in the failure of the entire bond.
DHCP should not be used.
The cluster would no longer communicate with the other node if the IP address changed.