Perform this task to create a cluster in UIS.
After you start creating a cluster, the system opens the cluster setup wizard page, and you cannot perform any other operations during cluster creation.
If a host to be added to a cluster is not in the same Layer 2 network as the management node, you must first configure management network settings for that host from its console and select scanning by IP address on the management platform. You cannot change the management network settings you have configured for that host.
To add a storage cluster, specify the management HA VIP of the storage cluster as its management IP. Then, UIS can manage the storage cluster at this IP.
If the system contains only one cluster, click Hosts on the top navigation bar, and then select Cluster Management > Create Cluster from the left navigation pane. If the system contains multiple clusters, click Hosts on the top navigation bar, and then select Create Cluster from the left navigation pane.
Click Start, and then click OK in the dialog box that opens.
On the UIS Setup Wizard page, configure the parameters as needed.
Review the configuration, and then click Finish.
UIS supports the virtualization and HCI deployment scenarios. The distributed storage component is deployed only in the HCI scenario.
To create an HCI cluster, select the HCI scenario. In this scenario, you must configure storage settings.
To create a compute cluster, select the virtualization scenario. In this scenario, you do not need to configure storage settings.
If you select the virtualization scenario during UIS initialization, you can select only the virtualization deployment scenario when you create a cluster.
Cluster Name: Specify a name for the cluster. The name cannot exceed 16 characters and must start with a letter or digit. Only letters, digits, and underscores (_) are allowed.
Host Naming: Set the host name prefix. The host name must be a string of a maximum of 32 characters that contains only letters, digits, and hyphens (-). It must start with a letter. The host name is automatically configured during the initialization process. After the initialization process finishes, you are not allowed to edit the host name. Plan the host name prefix in advance.
Numbering From: Specify the start number for host naming. A host name is a number attached to the specified host name prefix. The system assigns the start number to the management node and assigns the next available numbers to the service nodes joining the cluster afterward with an increment of 1. For example, if the host name prefix is cvknode and the start number is 1, the name of the management node is cvknode1 and the names of the service nodes joining the cluster are cvknode2, cvknode3, cvknode4, and so on.
Start IP: Configure the start IP addresses for the management network, storage front-end network, and storage back-end network.
By default, the management network configured for the management node is used as the management network. The management platform IP address must belong to the management network. The default start IP address for the management network is the start IP address of the management network. The system assigns the start IP address to the management network gateway by default. If you do not specify a management IP address for a service node, the system assigns the next available IP address to the service node with an increment of 1.
You can manually configure IP addresses for the management network, storage front-end network, and storage back-end network. Plan and reserve IP addresses for future cluster scale-up.
Storage front-end network IP addresses and storage back-end network IP addresses are used for internal access. You can use the default settings for the two networks.
VLAN ID: Select a VLAN ID for the management network. You must first configure VLAN settings from the Xconsole of each host. For more information, see management interface parameters configuration in H3C UIS Manager Installation Guide. You can configure VLAN IDs for the storage front-end network and storage back-end network.
Scan for Hosts: Select a scanning mode. If you select IP address scanning, the configured management network settings do not take effect.
Host Role: Select a host role, which can be management node, service node, or storage node.
Cluster Management IP: Specify the management IP address of the storage cluster.
Service Type: Select a storage service type, which can be block storage, file storage, or object storage.
For block storage:
Deployment Mode: Select a deployment mode for the disk pools automatically created by the system. Block storage supports the following deployment modes:
SSD caches+HDDs—The system automatically creates one disk pool. Deploy HDDs as data disks to store data and deploy SSDs as cache disks to accelerate reads and writes. This mode is commonly used.
All SSDs—The system automatically creates one disk pool. Deploy SSDs as data disks to store data without using data caches. Use this mode to provide high-performance storage services.
All HDDs—The system automatically creates one disk pool. Deploy HDDs as data disks to store data without using data caches. Use this mode to provide normal storage services.
HDDs+SSDs—The system automatically creates two disk pools. Deploy SSDs and HDDs as data disks in the high-performance disk pool and slow-performance disk pool, respectively, to provide storage services for applications that require different storage performance.
Provisioning: Provisioning mode of the disk pool.
Thick—Allocates a fixed amount of storage space to block devices in anticipation for the current and future storage demands. In this mode, you cannot allocate a capacity larger than the usable capacity of the disk pool to a block device.
Thin—Allocates storage space to block devices in a flexible on-demand manner. In this mode, you can allocate a capacity larger than the usable capacity of the disk pool to a block device.
Replicas: Specify the number of replicas for the block device automatically created by the system after initialization. The block device capacity is the smaller one of the following two values: 16 TB and the total effective capacity of the disk pool × 40%. For example, if the number of replicas is two and the total capacity of the disk pool is 20 TB, the total effective capacity of the disk pool is 10 TB. Because 10 TB × 40% = 4 TB < 16 TB, the block device capacity is 4 TB.
Cache Size: Set the cache size. This parameter is available only when you select the SSD Caches+HDDs deployment mode. The system divides SSDs into partitions based on the number of data disks and assigns a partition to each data disk as its cache. The larger the cache size, the better the performance. Make sure the cache size meets the following requirement: cache size × number of HDDs < total SSD cache capacity.
Once configured, the cache size cannot be edited. If more HDDs will be added in the future, make sure you have reserved sufficient SSD cache capacity.
Fault Domain: Fault domain type of the storage cluster. By using fault domains and a redundancy policy together, a storage cluster saves the replicas or blocks of data to different fault domains to ensure data security and high availability.
Rack—Each rack is a fault domain. Use this fault domain type when the cluster is large and contains a large number of racks.
Host—Each host is a fault domain. By default, the fault domain type is host.
Cache Protection Level: Select a cache protection level. This parameter is configurable when the deployment mode is SSD Caches+HDDs.
Standard—Cache and metadata are stored in cache disks in RAID0 mode on storage nodes.
Advanced—Cache and metadata are stored in cache disks in RAID1 mode on storage nodes. In the current software version, only some device models (such as R4900 G3) support the advanced level. For more information, see the compatibility matrixes.
For file storage:
Separating Data from Metadata: Select whether to save data and metadata separately.
Yes—Data and metadata are saved separately in two disk pools. Use this mode to provide high performance. If you select this mode, the system creates disk pools of the Object Storage-Data Pool type and Object Storage-Metadata Pool type automatically after initialization.
No—Data and metadata are saved in one disk pool. Use this mode in scenarios with limited storage resources. If you select this mode, the system creates a disk pool of the Object Storage type automatically after initialization.
Fault Domain: Fault domain type of the storage cluster. By using fault domains and a redundancy policy together, a storage cluster saves the replicas or blocks of data to different fault domains to ensure data security and high availability.
Rack—Each rack is a fault domain. Use this fault domain type when the cluster is large and contains a large number of racks.
Host—Each host is a fault domain. By default, the fault domain type is host.
Caching Mode: Select a caching mode for the system-created disk pool.
None—The disk pool does not have caches.
Buffer—You can create data disks and cache disks from the disk pool and use cache disks as the cache pool of the file storage data pool to accelerate data read and write.
Data Disk Type: Select the type of disks for storing data. Only disks of the specified type can be added to the system-created disk pool.
For object storage:
Separating Data from Metadata: Select whether to save data and metadata separately.
Yes—Data and metadata are saved separately in two disk pools. Use this mode to provide high performance. If you select this mode, the system creates disk pools of the Object Storage-Data Pool type and Object Storage-Metadata Pool type automatically after initialization.
No—Data and metadata are saved in one disk pool. Use this mode in scenarios with limited storage resources. If you select this mode, the system creates a disk pool of the Object Storage type automatically after initialization.
Fault Domain: Fault domain type of the storage cluster. By using fault domains and a redundancy policy together, a storage cluster saves the replicas or blocks of data to different fault domains to ensure data security and high availability.
Rack—Each rack is a fault domain. Use this fault domain type when the cluster is large and contains a large number of racks.
Host—Each host is a fault domain. By default, the fault domain type is host.
Deployment Mode: Select a deployment mode for the disk pools automatically created by the system. Block storage supports the following deployment modes:
SSD caches+HDDs—The system automatically creates one disk pool. Deploy HDDs as data disks to store data and deploy SSDs as cache disks to accelerate reads and writes. This mode is commonly used.
All SSDs—The system automatically creates one disk pool. Deploy SSDs as data disks to store data without using data caches. Use this mode to provide high-performance storage services.
All HDDs—The system automatically creates one disk pool. Deploy HDDs as data disks to store data without using data caches. Use this mode to provide normal storage services.
Cache Size: Set the cache size. This parameter is available only when you select the SSD Caches+HDDs deployment mode. The system divides SSDs into partitions based on the number of data disks and assigns a partition to each data disk as its cache.
Cache Protection Level: Select a cache protection level. This parameter is configurable when the deployment mode is SSD Caches+HDDs.
Standard—Cache and metadata are stored in cache disks in RAID0 mode on storage nodes.