The Struggle With IT Disasters
Businesses spend large amounts of money to secure server workloads against outages and data loss. The problem is that few disaster recovery (DR) plans sufficiently support VDI deployments, and downtime can be anything from power outages to malicious links and natural catastrophes. This can prevent employees from being able to work even if all server workloads remain online and functional. Furthermore, some organizations do not utilize a documented plan for disaster recovery either because they lack the necessary resources or are yet to recognize the importance of doing so. Its costs are often astronomical, ranging from hundreds to thousands of dollars that could otherwise be invested in operational strategies. Among companies that do have one, most account for infrastructure but fail to accommodate users who must remain productive at all times.
If you have come across physical desktop DR solutions, you probably know they are expensive and difficult to implement. Moreover, while many enterprises invest in VDI, SMBs have been unable to rationalize the move to virtual workspaces and make the most of the technology’s DR benefits. Fortunately, much of this has changed today as there is a growing awareness of the topic in question.
In this digital era, every business needs a robust platform capable of addressing both infrastructure and workspace recovery. Keeping this in mind, the first step is to evaluate budget, availability, applications users need access to, and factors such as keeping a DR plan up-to-date.
Let’s look at four key considerations when coming up with an IT DR recovery strategy. We shall also cover some measures one can take to quickly get virtual desktops up and running after a disaster strikes.
Setting up a project budget is the first and foremost concern for any organization and it plays an equally important role in an effective VDI recovery plan for disasters. There are many feasible approaches to this and each comes with its own price point.
For example, a company with adequate financial resources may build a thoroughly functional standby data center which it can utilize in the event of a disaster or tech failure. If it chooses not to incur the building and maintenance costs of a second data center, the IT department can rent space for its backup infrastructure in a colocation facility. In this scenario, a colo typically manages the building and power, cooling, bandwidth, and physical security features while the customer takes care of storage as well as servers. On the other hand, a business with a moderate IT DR plan budget may leverage a public cloud to host backup virtual desktops.
In any case, implementing VDI DR without running up massive costs calls for an in-depth quantitative strategy. There are two ways to achieve this.
The first option is to reduce risk by splitting VDI desktops and support infrastructure across two physical data centers. So, users will retain access to around half of the VDI sessions even if one data center becomes unavailable. In this case, the trick is to make sure that both data centers do not have shared dependencies. Moreover, even if IT chooses to build redundancy into a virtualization platform, their VDI instances should be capable of connecting to everyday applications. VDI instances also need to connect to authentication and management tools, such as Microsoft Active Directory (AD). This is a simple matter of understanding dependencies and testing regularly.
What if one only has a single data center, a small environment, and a limited budget? In this situation, using DaaS (Desktop as a Service), a cloud computing model to secure virtualized systems is a more feasible option especially when managing distributed users. Running Windows desktops from the cloud rather than a data center introduces the same benefits as local-network virtualization, including centralized desktop and data control and improved security. The only difference is that you need not pay the capital expense of a full-fledged VDI setup. What happens is you rent equipment from a cloud provider as part of your subscription cost.
For example, VMware offers a managed VDI service on multiple cloud platforms. Administrators can also use their own Azure Active Directory (AD) infrastructure to facilitate the underlying AD services which work with their cloud, on-premises, or hybrid environments.
Note that these cloud services currently cannot run both cloud-managed and on-site infrastructure. To tackle this, you must set up two different sets of infrastructure. While managing two sites may appear difficult, IT can quickly spool up new virtual desktops in the worst-case scenarios.
Alternatively, you can prepare a secondary VDI infrastructure if your company has a secondary network location. With hyper-converged infrastructure (HCI), it is possible to stand up an entire virtual infrastructure within rack space amounting to several units. This can be achieved without having to spend money on costly storage area network infrastructure. However, administrators cannot arrange DR configurations, so the solution is to engage with VMware to make the most of these deployments. Apart from this, IT personnel will have to perform other tasks, including configuring groups and managing the images.
For the best results, we recommend you to partner with a trusted vendor to streamline the budgeting and forecasting process.
IT personnel must determine how long an organization can manage a virtual desktop outage without affecting business operations and employee productivity to a great extent. Neglecting this aspect can negatively impact revenue, so the key is to ensure high availability at all times. Once again, budgeting is important because DR solutions that guarantee little or zero downtime are generally more expensive. Hence, the question: is there any affordable route you can take?
Let’s assume that your business utilizes the public cloud for disaster recovery, and IT creates backup virtual desktops in advance. Doing so allows you to stay prepared and power on instantly in the event that something goes wrong. However, cloud providers tend to bill customers based on resource usage, so a good practice is to have a hosting plan in place and generate virtual desktops only when needed.
Ask yourself about the applications users need access to, and understand the two types of apps you must consider in a DR plan for virtualization. Firstly, you have individually delivered virtualized applications, and secondly, there are apps installed onto virtual desktops directly. Keep in mind that an effective DR strategy enables users to access the applications they work with routinely, regardless of the scale and nature of a disaster.
In terms of application access, it is easy for companies to overlook two major factors: licensing and prioritizing time and mission-critical apps. Depending on the apps an enterprise uses and how IT delivers them, a group of standby virtual desktops can cause a spike in licensing expenses. Furthermore, some apps such as CAD/CAM/3D/Simulation and geospatial tend to be more business-critical than others.
When it comes to emergency situations, the DR plan should focus on bringing crucial resources online in the shortest time possible. Also, IT can simplify disaster recovery and decrease costs by eliminating any applications that an organization considers relatively unimportant from its DR plan.
IT should also prioritize the availability of user-profiles during an outage. Companies usually store team members’ profiles on a file server instead of within a virtual desktop itself. In such cases, administrators can opt for server or storage-level replication to build secondary copies of user profiles. For this to be successful, IT needs to ensure that the DR desktops can locate and connect to employees’ profiles.
Next, let’s study the concept of storage replication. A well-known solution in disaster recovery for virtualization involves setting up an infrastructure that spans numerous data centers or one which stretches to the cloud. In spanning data centers, administrators stretch the host cluster where the virtual desktops reside so that it extends across a number of data centers. The storage device that stores the virtual hard disks is duplicated to the secondary location, so copies of the virtual desktops are present in both places.
What if one considers failing virtualized desktops over to a secondary data center? Even with this possibility, it is better to create a different group of virtual desktops inside the secondary data center, as networking policies change when one hosts them in an alternate location. Another option is to connect employees to a virtualized desktop configured for use in the alternate location instead of reconfiguring current virtual desktops to function in an offsite area.
The importance of refining VDI DR plans can never be stressed enough, especially if we look at everything that is at stake. No single approach can consistently ensure smooth sailing, taking into account how quickly the tech landscape is evolving. The best way to work around this is to base your decision on employees’ job roles. For instance, think along the lines of a high-end DR plan for power users, and a less expensive option where task and knowledge users are concerned.
For further details about VDI solutions, please contact our support team. We will be happy to assist you with any queries you may have.