Search Docs...
⌘ K
Before expanding the cluster, make sure to apply to Arcfra in advance for the ACOS software license and the updated license code.
Refer to Hardware requirements to confirm that the nodes to be added meet the hardware configuration requirements.
Confirm the installed ACOS software version, and prepare the ACOS installation image files for nodes to be added in advance.
Confirm and download the VMware ESXi installation files and Arcfra VAAI-NAS addon installation files that match the version used by the cluster host.
For each new node to be added, plan four IP addresses: the SCVM management IP, SCVM storage IP, ESXi management IP, and ESXi storage IP. These IP addresses are used for the management and storage networks, respectively. Meanwhile, record the subnet mask, gateway, and other relevant information for both networks. If the existing ACOS integrates hardware management capabilities, you also need to plan IPMI IP addresses for the new nodes.
The IP addresses for a new node can be planned as follows:
The storage network of the new node must belong to the same subnet as that of the existing nodes in the ACOS (VMware ESXi) cluster.
The SCVM management IP and storage IP of the new node cannot be assigned to the same network segment.
The SCVM management IP of the new node must be assigned to the same network segment as that of the existing cluster nodes.
The SCVM storage IP of the new node can communicate properly with the SCVM storage IPs of the existing cluster nodes.
The SCVM storage IP and the ESXi storage IP of the new node must be able to communicate with each other, as well as the SCVM management IP and the ESXi management IP.
The IP addresses cannot overlap with any IP addresses within the following CIDR ranges:
If the new node previously belonged to another cluster, do not reuse its SCVM management IP address and storage IP address from the original cluster.
Information:
You can view the IP address configuration of existing ACOS (VMware ESXi) cluster nodes by logging in to AOC, selecting a host in the cluster from the left sidebar, and checking the host overview panel that appears on the right.
If the cluster to be expanded has RDMA or volume pinning enabled, make sure the software and hardware configurations of the node to be added meet the requirements for enabling these features as described in RDMA requirements and Volume pinning requirements.
If the cluster to be expanded has the NVMe passthrough hard disk hot-plug feature enabled, refer to Requirements for enabling hot-plug for NVMe passthrough hard disks for the relevant requirements. Ensure that the nodes to be added meet the requirements for enabling this feature and that the vmpctl file is prepared on the local host.
Ensure that the capacity specifications of the new nodes are consistent with those of the existing nodes in the cluster to be expanded.
If the ACOS cluster is all-flash and uses the same type of storage media, ensure that the new nodes are also all-flash and use the same type of storage media as the existing cluster.