Optimal use & tips

When using in a cluster use only SAS HDD and SSD. Avoid using SATA disks with additional interpozers.

Consider the number of columns when creating virtual disks. This can greatly affect performance. Custom number of columns can only be set when using PowerShell. Using more columns improves performance but limits the ability to add disks to the pool.

Example (Dual Parity virtual disk, 13 columns, 1000 Gb, Fixed Provisioning):

New-VirtualDisk -StoragePoolFriendlyName Pool1 -FriendlyName VD02 -ResiliencySettingName Parity -NumberOfColumns 13 -PhysicalDiskRedundancy 2 -Size 1000GB -ProvisioningType Fixed

Avoid using Parity and Dual Parity virtual disks when there is a large write load and/or random access by small blocks.

One of the most popular solutions today is a small 2-node cluster for Hyper-V. Due to the competitive handling of several VMs a rather high percentage of the load on the disk subsystem will consist of random access. When calculating the disk subsystem plan to use mirrored virtual disks (2-way mirror - standard double mirroring. For special scenarios you can use triple mirror). The performance of Parity and Dual Parity will be sufficient only for archive data. Plan on using at least 2 SSDs for tiering and write caching.

Avoid adding disks with different performance (for example, HDD 7200 and 10000 rpm) to the shared pool. Create virtual disks only from the HDDs with the same performance by selecting them manually in the GUI or via the -PhysicalDisksToUse parameter when creating through PowerShell. SSD disks must either be moved to a separate pool or used together with the HDD only as a fast tier and/or write-back cache.

< Previous | Content