Sizing Your Disaster Recovery Hardware Plan

sizing.JPGAs a Service Provider offering Disaster Recovery services,  you need to make sure that your customers replicate and set failover from their workload to your datacentre with as little as possible manual configuration for the best failover performance. If you are lucky enough and you are using

the Veeam Cloud Connect product, you know that the replication and networking configurations are automatic; that is a core function of the product in addition to the failover and fallback. Now let’s consider the sizing of the customer’s rental compute power? How you will size your Veeam hardware plans? In this post, I will discuss the ways you can size the Veeam Cloud Connect hardware plans before you offer your solution to your customers.

Before we start our Veeam Cloud Connect Hardware Plan sizing discussion, let’s first discuss several questions asked by some Service Providers, and then go through some possible answers.

What is the Veeam Cloud Connect Hardware Plan?

The Veeam Cloud Connect Hardware plan is simply a hypervisor resource, and in VMware, a resource group, which consists of CPU, Memory, Storage and Network connectivity. In other words, it’s a Veeam Cloud Connect product that multi-tenants the hypervisor, without using an additional product.

Do I need a Hardware Plan for each Tenant or Customer?

It depends on how you want to manage your offering. Veeam Cloud Connect allows you to share one hardware plan between many tenants; this then enables you to offer SLA types of hardware plans, or compute power type hardware plans. Alternatively, you can create a hardware plan running per tenant. Then you will be able to size for each tenant a hardware plan specific to the tenant’s workload necessary to deliver the compute power needed to be replicated to your datacentre. Both options are valid, but you have to ask yourself, as a Service Provider, about the setup overhead of each method, and how easy the tenant ongoing subscription manageability is going to be. I personally prefer configuring a hardware plan for each tenant. That way I can work with my tenants to size the Hardware Plan, and also ensure I never over, or undersize, the tenant workload reservation.  This will save me money on my hypervisor usability. Also, a Hardware Plan for each tenant will make it easier to manage and monitor my hypervisor environment.

How can I determine the tenant hardware plan size?

In this blog post, and to simplify the explanation, I will provide an example for sizing a hardware plan which will host only one replica virtual machine. So let’s started.

As I mentioned before, the Veeam Hardware Plan, a resource pool when in VMware, defines CPU, memory, storage and network. After a hardware plan is assigned to a tenant, the tenant can consume these resources in the form of virtual infrastructure. The virtual infrastructure is available to replicate the desired workload to the Service Provider and to provide for the initial failover in the event of a disaster.

The figure below shows how Veeam Cloud Connect can configure the hardware plan compute, changing the Gigahertz on the CPU, and the Gigabytes in the memory.

Veeam Cloud Connect Hardware Plan – CPU and Mem Config

The question I have been asked by Service Providers is this, “How can I size the CPU?” For example, my tenant has two CPU cores for the workload he wishes to replicate to the Disaster Recovery site. The answer is simple, think about the GHz the tenant needs for his CPU; to get the accurate GHz for any workload, Service Providers can use VeeamOne to extract this information. The process is reasonably straightforward and involves running the VeeamOne product to determine the workload maximum, and average CPU consumption, over a period of time.

For example, I monitored my server that I wish to replicate to a Service Provider’s environment for a day. While monitoring the server, I realised that the Max CPU power the server is consuming is 1.15GHz and the average is 0.11 GHz.

VeeamOne – CPU maximum and average usage

The memory also showed maximum usage to be 1.92 GB, with an average usage of 0.27 GB, and memory consumption of 3.98 GB average.

VeeamOne – MEmory maximum and average usage

Now that I know the CPU GHz, and the memory GB from the VeeamOne tool, I decided to create the tenant Hardware Plan with 1.5 GHz and 4 GB RAM.

Applying the configuration on Veeam Cloud Connect


The Veeam Cloud Connect product is a piece of very sophisticated software. It offers many configuration features to allow you to simplify your remote Disaster Recovery deployment for your organisation, and for customers seeking to implement a Disaster Recovery solution quickly and cheaply for their critical workloads.

On the other side, Service Providers using the Veeam Cloud Connect product can simply and inexpensively multi-tenant their virtual infrastructure to host multiple tenants, their customers, securely. They can do this without the complexity and high cost associated with other products on the market.

I hope this blog post provided some clarity on sizing your Veeam Hardware Plan to satisfy your tenants, and in the process, reduced your datacentre infrastructure costs.

Leave a Reply