{{ $t('productDocDetail.updateTime') }}: 2026-01-05

Description

HCI supports multiple volume management. Common volumes and stretched volumes can exist in a cluster at the same time. Multiple virtual datastore can be created in a HCI cluster. Multi-volume can be formed by host or disk. Units and volumes are physically and logically isolated, but for virtual machines, volumes can be selected and migrated, and they can also meet the heterogeneous hardware requirements of different services in the cluster. The following operations are the procedure for adding normal volumes.

A host-based multi-volume requires at least six hosts, three hosts for a volume, and the other three for a volume. Three or more hosts can divide multiple volumes for multi-volume based on disks. Therefore, a single host can create an all-flash volume based on creating a mixed volume, which will require high disk performance. All services run on all-flash volumes to improve disk usage. Host multi-volume and disk multi-volume methods can coexist on an HCI cluster.

As an example, the supported group multi-volume mode for three hosts is shown in the figure.

Precautions

1. The newly added volume will be formatted when it is created. Before adding, please make sure that there is no data that needs to be preserved on the disk used to add the new volume.

2. A cluster is allowed to have a maximum of six virtual datastores, of which a maximum of one stretched volume is allowed.
3. A volume consists of hard disks on at least three hosts, and a host can be divided into two volumes at most.
4. Virtual storage does not support a single copy. When a single copy is used, the virtual storage is not redundant, and there is a risk of data loss.
5. Two hosts do not support data balancing.
6. Striping is not supported for virtual machines on two hosts.
7. Virtual storage is a full HDD deployment and does not support upgrades or new deployments.
8. If the capacity of three hosts is too large, the deployment of virtual storage will be limited. It is required that the capacity of one host cannot exceed the sum of the capacities of the other hosts. For example, host A is 1TB, host B is 1TB, and host C is 5TB. Deployment is not supported.
9. A host spanned by one volume cannot belong to two other volumes.

10. Stretched clusters do not support multiple volumes. As shown in the figure below, a host cannot belong to both extended and normal volumes.

11. All-flash volumes (compression) do not support hard disk volumes but only host volumes.
12. A single host in an ordinary volume has at least one SSD, and a single host in an all-flash volume has at least two SSDs.

Prerequisites

1. To set up multiple volumes, each volume requires at least three hosts.

2. At least three hosts are required to form multiple disk volumes.
3. The expansion host's management interface, storage interface, overlay network interface, and service interface have been connected to the network.

Steps

Creating Storage Volume

1. Navigate to Nodes and click Add New Node.

2. Select the node to be added. If the node to be added is not in the list, click + to add it.

3. Adding a host will delete the firewall configuration of the added host. Click OK.
4. If it is detected that the MTU of the overlay network interface is inconsistent, reconfigure it, and check the Enable high-performance mode checkbox.
5. After adding a node successfully, configure the communication interface. Then, the new node will be visible under Home.

6. Under Storage > Virtual Storage > Virtual Datastore, click New to create a new virtual datastore.

7. Click Settings to configure storage interface and IP address.

8. Configure the deployment mode according to the Topology as shown below.
9. Select the storage interface and configure IP address.
10. Click New again to create new Virtual Datastore and select Next.

11. For Method, select Use unused disks added to existing datastores.
12. Use disks on new hosts: Select three or more hosts in the cluster to create a new volume.
13. Use unused disks added to existing datastores: Select the unused hard disk in the host contained in the volume and choose to create a new volume.

14. Select the disks that are required for the capacity expansion.

15. Configure the disk groups for the new datastore. If it is not a full flash deployment, SSD is required for use as a cache disk.

16. Verify the configurations and complete the procedure.

Adding hosts

1. Click Add New Node in Nodes.

2. Select the host that needs to be added. If the host is not shown here, click + icon to add manually.

3. When adding a new host, it will remove the firewall configurations of the host. Click OK to continue.
4. If it is detected that the MTU of the communication interface is inconsistent, reconfigure it and check the Enable high performance mode checkbox.
5. 
6. After adding a node successfully, configure the communication interface. Then, the new node will be visible under Home.

7. Navigate to Storage > Virtual Storage and click New.

8. Click Storage Area Network Settings, and it will redirect to Communication Interface > Storage Network Interface. Select Reset Storage Area Network to configure the storage interface and IP address.

9. Configure the deployment mode according to the Topology as shown below.

10. Select and configure the storage interface and IP address.

11. Click New again to create new virtual datastore and click Next.

12. For the Method, choose Use unused disk added to existing datastores.
13. Use disks on new hosts: Reselect 3 or more hosts in the cluster to create a new volume.
14. Use unused disk added to existing datastore: Select an unused hard disk in the host that the volume contains to choose to create a new volume based on the existing volume.

15. Select the disks that needs to be included in the disk group.

16. For the newly created disk groups, an SSD as a cache disk is necessary if it is not a full flash deployment.

17. Verify the configurations, and click OK to complete the operation.
18. Datastore type configuration: Select Storage > Virtual Storage and click New. Select the Type as Ordinary Datastore.

19. Select Node: Select the nodes from the cluster, and choose the nodes that need to be included in the virtual datastore volume.

20. Select the hard disks that need to be added. Configure the hard disk and disk groups, then set up the use plan for each disk on the hosts. The system will automatically detect the disks of all hosts in the cluster. By default, hard disks are selected as data disks, and solid-state disks are used as cache disks. Using the default configuration is recommended. If you need to deploy multiple disk volumes, you have to plan for backup disks on the second volume group.

21. Verify the configurations, and click OK to complete the process.

22. The following page displays virtual storage configurations, including available disk space, number of data copies, and the total number of disks. After confirming configurations, click OK.  Then, input the administrator account password: admin. Then, click Finish to begin initializing virtual storage.

Creating Storage Multi Volume

1. Click Add New Node in Nodes.

2. Select the node to be added. If the node to be added is not in the list, click + to add it.

3. Adding a host will delete the firewall configuration of the added host. Click OK.
4. If it is detected that the MTU of the overlay network interface is inconsistent, reconfigure it, and check the Enable high-performance mode checkbox.
5. After adding a node successfully, configure the communication interface. Then, the new node will be visible under Home.

6. Go to Storage > Virtual Storage > Virtual Datastore and click New.

7. Click Settings to configure the storage interface and IP address.

8. Configure the deployment mode according to the Topology as shown below.

9. Select the storage interface and configure IP address.

10. Click New again to create a new Virtual Datastore and select Next.

11. For the Method, choose Use unused disk added to existing datastores.
12. Use disks on new hosts: Reselect 3 or more hosts in the cluster to create a new volume.
13. Use unused disk added to existing datastore: Select an unused hard disk in the host that the volume contains to choose to create a new volume based on the existing volume.

14. A list displays node information such as node name, node IP, total SSDs, and HDDs. You should select the node(s) from the list you want to add to the virtual datastore. Note that at least three nodes are required to create a second virtual datastore.

15. Specify the use of the disks. Disks added to virtual storage can be used as the data disk, cache disk, or spare disk. If it is not a full SSD, it is necessary to have an SSD as a cache disk.

16. Verify the configuration and complete the operation.
17. Navigate to Storage > Virtual Datastore, click New, and choose the Type as Ordinary Datastore.

18. Select Use unused disks added to existing datastores and select the nodes needed to add to the virtual datastore.
19. Select the hard disks that need to be added. Configure the hard disk and disk groups, then set up the use plan for each disk on the hosts. The system will automatically detect the disks of all hosts in the cluster. By default, hard disks are selected as data disks, and solid-state disks are used as cache disks. Using the default configuration is recommended. If you need to deploy multiple disk volumes, you have to plan for backup disks on the second volume group.

20. Specify the use of the disks. Disks added to virtual storage can be used as the data disk, cache disk, or spare disk. If it is not a full SSD, it is necessary to have an SSD as a cache disk.

21. Verify the configurations.

22. The following page displays virtual storage configurations, including available disk space, number of data copies, and the total number of disks. After confirming configurations, click OK. Then, input the administrator account password: admin. Then, click Finish to begin initializing virtual storage.