Disaster recovery as a service (DRaaS) is a hot topic and an exciting service for cloud service providers. But...
before there can be a discussion on what to look for from DRaaS providers, there has to be an understanding of the services they offer.
In reality, DRaaS is a very broad spectrum of potential services. The technology solves the problem of having another location to bring up an organization's entire IT in the event of disaster. It eliminates duplicate hardware and software requirements, and has the potential to save a company a lot of money.
Here are some services that can be offered by DRaaS providers:
- Simple data recovery for administrators and selective recovery from within the service provider's data center.
- Virtual machine mounting (and physical-to-VM conversion) within the service provider's data center for temporary application availability.
- Full disaster recovery within the service provider's data center, including optional services such as:
- VM mounting (and physical-to-VM conversion)
- VPN rerouting
- DNS updating
- User access via virtual desktop infrastructure (VDI)
- Physical access for administrators
- Ability to run operations for extended periods of time
- Professional, skilled, experienced staff to facilitate disaster recoveries
- Annual, bi-annual or quarterly testing
- DR planning help
The key to selecting a DRaaS provider is to ensure they deliver what your organization requires to stay up and running. Don't assume a service provider offering DRaaS will cover all disasters and deliver all possible services. The worst time to discover that the service provider cannot provide a required service is during a disaster.
Conduct an internal DR assessment
Before you evaluate DRaaS providers, you should conduct an internal assessment as to what is required organizationally by applications, users and customers. In other words, what does disaster recovery mean to the organization?
Does it mean zero downtime and zero loss of data, or some defined period of downtime and a loss of some data? How does this requirement vary by application, workload and so on? And what types of disasters are you protecting against (hardware failures, human errors, malware, natural disasters, terrorism)?
Finally, what is the budget? The old rule of 90/10 applies here. The last 10% of RTO and RPO often represents as much as 90% of the cost. The closer either or both get to zero, the more rapidly costs go up. Determining the answers to the above questions is crucial in establishing whether a DRaaS provider can meet your requirements.
Essential questions to ask DRaaS providers
Many DRaaS providers are new to the DR and business continuity market. This is because DRaaS is a natural outgrowth of the backup as a service market, and cloud backup is far simpler to deliver. Picking a service provider necessitates a thorough vetting to ensure the services purchased are deliverable and meet expectations. Do not expect a service-level agreement (SLA) to guarantee that expectations are met. SLAs typically provide a refund of payment if the service provider fails to contractually deliver. The key is asking DRaaS providers the following questions before selecting a service.
- What percentage of the service provider's customers can be supported concurrently during a regional disaster such as a hurricane? What DR resources are available for recovery and how many customers can be supported concurrently? How are these resources managed, tracked and updated? What happens if the provider cannot provide DR services?
- What are the rules for declaring a disaster? Is it first-come, first-served until resources are maxed out? What happens to those who cannot be serviced?
- How will users access internal applications? Will VPNs be managed/rerouted? How does VDI affect user access and who manages it during disaster recovery?
- How will customers, partners and users access outward-facing applications? Will DNS nodes be updated for outward or customer-facing applications?
- How do you ensure administrators and users receive access to servers/applications?
- What are the procedures for failback?
- What professional services/skills/experiences are available from the service provider to facilitate disaster recovery and how much do they cost? How much help can be expected in a DR event?
- What are the DRaaS provider's testing processes? Can I perform my own testing?
- How long can a customer run in the service provider's data center after a disaster is declared?
- What are the costs associated with the various DRaaS options? Are they a la carte, bundled or priced upfront? Is there a mix of up-front and recovery event costs?
Disaster recovery is a non-trivial event. DRaaS has the potential to be a cost-effective, efficient DR/BC solution, but only if it meets your organization's requirements.
Deciding between an IaaS host or MSP for disaster recovery
When premium disaster recovery service providers make sense
Why DRaaS makes sense for SMBs