Blog & Events

Does Software-Defined Data Center as a Service provide IT resiliency?

Apr 5, 2018, 11:00 AM by Trenton Baker

Data Center as a Service

“Software-defined data center” (SDDC) started life as a VMware umbrella marketing term for a collection of software-defined technologies. Since then it’s become an industry-standard term, and Microsoft has adopted it for some Hyper-V solution sets.

Today some enterprising vendors are offering SDDC as a cloud service (SDDCaaS). The offering is not radically different from Infrastructure as a Service (IaaS); there are some technical distinctions, but functionally the offerings are similar. Your IaaS service may even run on your cloud provider’s SDDC.

Yet no matter how you define your cloud-based infrastructure service, an SDDC is only as good as the applications and data running on it. If your infrastructure blows up and your applications go down with it, you need to have application and data protection in place for immediate failover or restore.

Don’t short data protection and disaster recovery out of a false sense of security, or because you’re burning through your SDDC as a Service budget. If you are storing active data in the cloud, then you need to protect it. And to protect it, you are going to need Backup as a Service and Disaster Recovery as a Service.

Defining the Software-Defined Data Center

Let’s start by explaining the software-defined data center. An SDDC is a computing infrastructure that extends virtualization services like resource pooling, logical abstraction, and automation to data center services. Critical components include the enabling hypervisor, software-defined networks and storage, and a management layer.

SDDC does not require a cloud to run, living happily on its own as an on-premises infrastructure. There are three basic categories for integrating SDDC with a cloud: as a private cloud infrastructure, in a hybrid model where an on-premises SDDC extends operations to the cloud, and as a service where a cloud provider delivers SDDC functionality from a public cloud.

  1. Private clouds and on-premises SDDC are a little tricky. Some people believe the two are synonymous, but they’re not. An SDDC might provide the private cloud’s infrastructure, but a private cloud is more than the infrastructure layer and does not have to be software-defined.
  2. Hybrid clouds and on-premises SDDC extend some private cloud operations to the cloud such as cloud bursting or backup and restore.
  3. SDDC as a Service is delivered by a cloud provider to end-users. In this model, the SDDC provider creates the SDDC in the cloud, automates processes, adds end-user management tools and resources, and delivers the virtual infrastructure to its cloud customers.

Option #3: Software-Defined Data Center as a Service (SDDCaaS)

Let’s nail down the “as-a-service” term first. In 2015, Gartner defined SDDC as “a data center in which all the infrastructure is virtualized and delivered ‘as-a-service.’” Today the “as-a-service” term better describes a cloud computing subscription model delivered from the cloud to end-users. That’s how we’re using the term to describe SDDC as a service.

Does the concept of delivering data center infrastructure as a service sound familiar? It should. It is like Infrastructure as a Service (IaaS), which provides the infrastructure computing stack of servers, network, storage, and virtualization services. From that point, the user deploys middleware, operating systems, and applications. So, what’s the difference?

As with private clouds, there is no defining requirement that IaaS must be a software-defined data center. The IaaS provider manages the underlying infrastructure including servers, network, storage, and virtualization services -- but the data center infrastructure may or may not be software-defined.

The IaaS environment may, in fact, be an SDDC —VMware refers to its SDDC products as IaaS with automation and self-service. But there is no essential requirement that it be one.

Is SDDC the same as the Cloud?

A significant architectural change began with the concept of the Software-Defined Data Center as the separation of the controlling data planes and the introduction of elasticity and automation.

While VMware and others have touted SDDC architecture as a revolutionary building block for Infrastructure-as-a-Service (IaaS), some have mistakenly assumed that SDDC is now synonymous with Cloud Computing. The common understanding of SDDC is not akin to the Cloud but rather an enabling technology to help build a holistic ‘as-a-Service’ cloud solution or as proposed here, SDDCaaS.

The SDDC vision is to not merely virtualize the underlying hardware infrastructure but to allow it to dynamically and automatically provision from local virtualized resources up through to the cloud. Data protection and continuous availability can be achieved with cloud bursting for dynamic capacity or spinning up new VMs in high demand instances, to VDI boot storms, good ol’fashioned scheduled maintenance and yes even providing IT resiliency in disaster recovery scenarios.

Protecting Data in SDDCaaS

However, the crucial point is not whether you define your cloud-based data center as SDDCaaS or IaaS. It’s this:

Whether you have SDDCaaS or IaaS, you had better protect your data.
Protect the data on your cloud data center with Backup as a Service (BaaS), and Disaster Recovery as a Service (DRaaS).

Backup as a Service (BaaS)

BaaS is long-term backup that is delivered as a service to your SDDCaaS. Most megacloud software-defined data center providers may backup the data center’s active data but their SLAs state that they will not backup data that is older than 30 days. The cloud provider may offer extended backup and restore functionality, but you will pay for it as a separate service. You should also consider third-party BaaS providers. For example, Veeam Cloud Connect protects data in IaaS by copying your data to your on-premises data center or to another cloud data center.

BaaS should be able to backup and restore data running on your cloud software-defined data center and its workloads: VMs, SDDC hypervisors, virtual networking and storage, monitoring and management services, user login credentials, and more. On top of that, BaaS should operate cloud-native backup and restore for file, object, and block backups – including snapshots, replication, or versioning.

Your BaaS provider will work with you to build long-term data retention and restore strategy that meets your RTO/RPO. Many BaaS providers also include archival services and searchable indexes.

Disaster Recovery as a Service (DRaaS)

DRaaS is online disaster recovery that keeps critical applications available during downtime. Its most familiar use case is replicating on-premises virtual environments to the cloud for immediate failover processing. However, with cloud usage expanding so rapidly you may need to consider DRaaS for your cloud SDDC service.

Understand what your SDDCaaS provider offers for disaster recovery and if their offering is a basic or premium service. At the minimum, you need to understand their cloud-native DR capabilities for replicating your SDDC workloads and failover/failback, and their RTO and RPO ranges. Also ask if their DR functions include only the software-defined data center failover, or if it as well replicates and fails over middleware, OS, application, and data layers running on top of it. Ask the same questions of third-party DRaaS providers.

SDDC as a Service gives you the advantages of a software-defined data center without the headache of building your own. But you are ultimately responsible for your application availability and the safety of your data. Adopt BaaS and DRaaS to protect your software-defined data center and the valuable information it supports. And your peace of mind isn’t a bad reason either.


Readers of this blog post are also interested in the webinar:

The Role of the MSP in the Multi-Cloud

Get Your FREE
Market Analysis Brief!


Disaster Recovery Planning

Download a free Planning Guide

“Storage Switzerland details DR Planning from Good to Great”