June 28, 2020
Topics: Cloud Tiering Data TieringFile ServicesAdvanced6 minute readVDI
Virtual desktop infrastructure, or VDI, is a beneficial technology. In a VDI, users throughout an organization on low-powered, inexpensive personal laptops and desktops connect to and use the resources of high-performance servers. The typically varied user workloads shared across servers require less overall CPU and therefore reduced power costs while keeping the generated heat and noise solely in the data center.
With a VDI, users don't need to be in the office, or even use company-issued computers, which makes this technology extremely useful for many companies given the ongoing workplace disruptions. Users can work safely from home, but as if they were in the office and the data is secure because it never leaves the VDI; the users' machines never directly access the company network, which helps prevent or stop malware rampaging through the corporate network.
However, over time, like any desktop, cold data accumulates in a VDI deployment. Unused files, volume snapshots, DR environments, and offsite backups are all cold data that reduce the effective capacity of your performant storage.
In this article we’ll take a closer look at VDI deployment and explain how NetApp’s Cloud Tiering service can reduce the cold data on your performance tier, increasing available capacity, and reducing potential CAPEX for increasing capacity for your VDI.
Virtual desktop infrastructure is a technology where each user's desktop runs on a virtual machine hosted on a hypervisor. The hypervisor may be part of the same virtual infrastructure that hosts server VM's or specially built to host virtual desktops.
Each user has a thin client, usually a low spec laptop or desktop, which is used to connect to the virtual desktop. The VDI provides the opportunity to reuse older laptop and desktop machines that are not capable, instead of buying new, perhaps as part of an environmental initiative, as many enterprises are seeking to go green.
The user may connect to one of a group of identical virtual desktops or have a specific virtual desktop reserved for them, which they can personalize, depending on the infrastructure's configuration. With appropriate VPN software, the user doesn't need to be in the office to access the virtual desktops; in fact, as long as network connectivity exists, the virtual desktops are accessible from anywhere. The VPN client connects to the VDI network, and the VDI client accesses the virtual desktops.
If a thin client is lost, stolen, or infected with a virus or trojan, the data is secure because VDI isolates the user's machine from the company network. Reducing the incidence and spread of viruses, and the data loss caused by trojans. But if infected, a virtual desktop can be quickly deleted and a new one provisioned in minutes.
Provisioning a virtual desktop involves cloning a purpose-built desktop VM template into a new virtual desktop and can be automated. VDI technology can auto-provision new virtual desktops when the number of spare virtual desktops falls below a specified value to ensure there are always available virtual desktops.
How NetApp ONTAP Benefits VDI
ONTAP has many functions that can optimize your virtualized environment, providing an improved end-user experience and a cost-effective solution for IT with All-Flash FAS arrays or SSD-backed FAS arrays, write optimizations, and storage efficiencies.
Using ONTAP increases business agility, reduces scheduled downtime, and simplifies storage management. For example, with a volume move operation, IT can seamlessly and non-disruptively move entire pools of users or pods from one node to another within a storage cluster. This operation can be used for maintenance activities, hardware upgrades, or even load balancing. And with FlexClone® technology, which creates instantaneous, space-efficient clones of virtual disks, IT can clone VMs in a fraction of the time it would take with the native hypervisor tools, by essentially offloading the task to ONTAP. Storage efficiency features, such as deduplication, compression, and compaction, help IT departments significantly reduce capital costs. With highly efficient NetApp Snapshot™ technology and the ability to replicate them offsite, and using SnapMirror® replication, IT can easily and efficiently implement their data protection strategy protecting the important data both locally and remotely.
However, in virtualized desktops environments, downloaded files and media can become old quickly, as will unused data blocks of the operating system and applications. Even the very efficient NetApp Snapshot copies are considered cold data since they’re rarely used. When they’re occupying space on the high-performance storage tier, they can reduce the available capacity, and, therefore, the number of virtual desktops your VDI will have available. The same is true for disaster recovery environments and offsite backups that contain copies of the entire dataset of the VDI, which are usually on performant storage so recovery can take place as fast as possible.
A Cold Data Solution for VDI: Cloud Tiering
If you are deploying a VDI with a NetApp AFF or SSD-backed FAS system, you have a powerful built-in resource at your disposal that can help reduce your overall costs and reclaim storage space on your performance system: the Cloud Tiering service. With Cloud Tiering, your cold data is migrated to public cloud-based object storage on AWS, Azure, or Google Cloud, or to an on-premises NetApp StorageGrid, recovering significant capacity on your storage performance tier.
Whenever Cloud Tiering classifies a data block as cold, it moves that block to cloud storage, based on the tiering policy you set for each volume. Once configured, data tiering is automatic and seamless, requiring zero changes at the VDI layer. Plus, storage admins can optimize the cooling period between 2 and 63 days.
Once marked as cold, the blocks are migrated to cloud storage, freeing up considerable space on the performance tier. When the data stored in the cloud is accessed by random read operations, it is rewarmed and brought back to the on-premises performance tier (note that this operation won’t be triggered by sequential reads, such as virus scans, when using the Auto tiering policy).
Cloud Tiering offers three tiering policies. The Auto cloud tiering policy is an excellent choice for virtual desktop volumes. Under the Auto tiering policy, user and snapshots data blocks are considered cold data if they are not accessed for the cooling period.
The Snapshot-only policy marks only cold snapshot blocks for tiering. Since snapshot copies may need to be retained locally for compliance with regulations such as GDPR or HIPAA or company data protection policies, the Snapshot-only policy can wind up recovering a significant amount of the performance tier’s capacity.
The All tiering policy moves entire volumes to the cloud, which can be very beneficial for off-site backups, and disaster recovery copies. There is no cooling period, and if accessed, the blocks are read from the cloud tier and never moved back to the performance tier. This policy recovers almost all the space of the volume.
During a problem that requires failing over to the DR environment, the tiering policy should be changed to either Auto or Snapshot-only. Otherwise, all data is read from the cloud tier, and performance is affected.
More Room for VDI Performance
With VDIs, companies gain a powerful method for enabling a dispersed workforce; however, as with any IT deployment, space can become scarce, forcing unexpected expenditures. To limit the increased CAPEX expenses, Cloud Tiering can recover capacity on the storage performance tier, to store more VDI VM's and their associated data or reduce the required capacity for VDI. For new infrastructure design, this allows the architect to specify a highly efficient lower-capacity solution, reducing CAPEX.