Welcome to the cloud, home to numerous as-a-service solutions. From CRM to computing, businesses are buying these services in an attempt to improve the core aspects of their organizations. But one cloud service stands out as misunderstood and under-evaluated, yet incredibly useful and valuable: Disaster Recovery as a Service (DRaaS).
Disaster Recovery as a Service (DRaaS) is a cloud computing and backup service model that uses cloud resources to protect applications and data from disruption caused by disaster. It enables the full replication and backup of all cloud data and applications, while serving as a secondary infrastructure. This secondary infrastructure establishes a new environment that allows an organization and its users to continue with daily business processes while the primary system is repaired. DRaaS allows applications to run on virtual machines (VMs) at cloud disaster recovery centers at any time and has additional capabilities to accommodate on premise infrastructure by replicating primary data center systems at cloud and colocation recovery centers.
While DRaaS is quickly gaining in popularity, some companies are still left in the dark when it comes to this important, evolving service, especially in always-competitive business climates.
A major reason for the explosion of DRaaS offerings lies in the general evolution of recovery services. Simply put, the 20th-century version of disaster recovery is outdated. Businesses no longer need to rely on antiquated “cold site” methods of moving tapes to off-site DR systems to support data recovery. Now, with the advent of virtualization and the ability to instantly replicate data with Continuous Data Protection (CDP) technologies, disaster recovery is completely different. It is now near real-time, allowing businesses to quickly copy and send data across a network and into hosted, virtualized environments.
Hello affordable solutions
DRaaS is also catching on as businesses realize the need to improve their recovery objectives without increasing their DR spend; DRaaS provides faster recovery solutions at a lower cost than traditional recovery services. Given the evolution of technology and public cloud services, any organization can afford to backup and recover its data swiftly in replicated storage and compute environments; while in the past, only the most elite of corporations could achieve such a goal at a cost that didn’t consume the whole IT budget. Although larger organizations with primary / backup data centers have often run SAN-to-SAN replication for their storage environments without reliance on traditional tape-based systems, with DRaaS, smaller organization can now achieve the same capability without the need to purchase a traditional tape backup environment and its associated hardware.
Therefore, DRaaS solutions are the de facto choice for data recovery, and businesses agree. According to a study from TechNavio, DRaaS is one of the fastest growing cloud-based segments in the Infrastructure-as-a-Service (IaaS) market, and adoption rates are set to double year over year.
New choices beyond traditional recovery
At the end of the day and with the right perspective, DRaaS seems like an obvious choice. After all, simply storing and backing up data with Amazon Web Services cloud storage solutions, or on Azure and its storage systems, can become costly and complex. DRaaS providers are different than AWS and Azure in that they provide customers with a way not only to vault their data and applications in the cloud, but also help recover their environment to hardware, virtual machines or another cloud environment, if needed. DRaaS providers have the additional benefit of offering the expertise to orchestrate the recovery by doing things like application recovery verification. DRaaS providers can also give companies specific service levels based on the applications and data that need backup and recovery. In addition, DRaaS providers have global Cloud-Colocation Centers with the ability to standup / co-locate legacy infrastructure. DRaaS providers can offer traditional disk or tape-based storage along with cloud recovery services, from either the same or different recovery center locations depending on customer diversity needs.
DRaaS is essentially providing an alternative, pay-as-you-go infrastructure with flexible contract models (some DRaaS providers don’t require time-bound contracts). This allows customers to customize their DR service to accommodate their changing technology environment and business needs.
So, it’s time for businesses to prepare for DRaaS procurement, and there are three steps to that process: First is the evaluation of internal systems, then comes evaluating DRaaS providers and lastly, companies must create their recovery strategy in the context of using DRaaS capabilities.
Determine disaster recovery functional requirements
The first step in preparing to evaluate DRaaS services is to perform a Business Impact Analysis (BIA) and/or a Technology Impact Assessment (TIA). These exercises define the most critical business processes and which are most impacted by application downtime. The analysis provides the basis for disaster recovery functional requirements and a model in which applications are prioritized and organized into tiers, based on their sensitivity to downtime. Each tier can then be defined by the recovery time (RTO) and recovery point (RPO) requirements of the applications within it.
Presenting this information, along with an inventory of current virtualized, physical and legacy infrastructure to DRaaS providers will allow them to apply the data movement technologies in their service / product portfolio for compute, storage and vaulting to the application tiers and the respective RTO / RPO objectives defined in the BIA / TIA.
Recovering from system failure is difficult and time-consuming without modern disaster recovery solutions.
Sizing up DRaaS provider capabilities
After identifying their disaster recovery requirements, businesses can engage DRaaS providers so they can begin to assess their breadth of services, the platforms they support and their proven successes with other customers. Here are a few considerations:
- Does the DRaaS provider offer prepackaged services for DR failover to a cloud environment including pay-per-use solutions with varying rates depending on RPO and RTO targets?
- Can the company run production environments out of the cloud during disaster declarations or testing? What are the extensible DR testing capabilities?
- Depending on their cloud recovery offering, how many tests per year do they offer before charging to use compute capacity and what are those costs? Do they charge a Disaster Declaration Fee?
- Does a provider offer native SAN-based replication services? Can they integrate their services with your current storage environment?
- What Continuous Data Protection (CDP) technologies does the DRaaS provider employ in their cloud-based recovery solutions?
- Is there self-service management available via a centralized portal (e.g. integration of continuous data protection technologies into their enterprise cloud services portal)?
- What are the solutions for public / private cloud replication with burstable usage for V2V, P2V, etc.?
- Will the provider virtualize existing network appliances or provide equivalent solutions?
- Does the provider co-locate infrastructure (e.g. backup appliances, tape library, legacy systems, like Solaris, Oracle DB, etc.) in their own data centers or do they use a partner organization to do this?
- Will the provider offer managed vaulting services?
- What is the DRaaS provider licensing model on legacy systems that need to be co-located?
- Are there a wide range of network connectivity options to a provider’s data centers, including accommodating connectivity to the co-lo space via 10Gb LAN connections?
Can the DRaaS vendor provide a schedule of SLAs and rate card tied to their cloud-based and traditional recovery service offerings?
Most customers are going to have a mixed environment of old and newer technology, and often there are legacy applications that cannot be virtualized. Good DRaaS providers offer the ability to accommodate a hybrid environment of both virtualized X86 as well as legacy compute platforms, if needed.
Solution design considerations
Different DRaaS solutions and data centers call for different strategies, so businesses should not forget to create a specific recovery plan or “DR Playbook” for their continuity, technology and recovery needs. This DR Playbook will be a driver for the solution design. Here are some strategic solution design considerations:
- Alternate approaches / DR technologies: For example, cloud backup file level, VM disk level and VM Image restoration capabilities negate the need for traditional physical equipment, e.g. tape drives.
- Virtual tape libraries at DR site: Are virtual tapes required and where will they be hosted?
- Consolidation strategies: Customers could potentially reduce costs of maintaining their own network appliances using DRaaS-provided virtual instances of these appliances within the context of the DR service. This includes network security devices like firewalls and other network-based appliances like load balancers.
- Active-active environments vs. virtual disaster recovery at a “hot site”: Assess the cost-benefit implications of running near real-time data replication and recovery because active-active often requires a customized design effort.
- Synchronously or asynchronously: IT teams need to determine whether target volumes should be kept updated in real-time or asynchronously; dependencies include current array hardware and network latency .
- Costs: What are businesses paying per vCPU/GB, vRAM/TB storage space for per-resource cost estimates?
- Virtual environment structures: Are there any hypervisor-protected VMs using RDMs or clustered disks in the current environment?
- DR testing using production systems vs. establishing test-specific “bubble” network: DRaaS can provide DR tests running in a sandbox environment without affecting production systems
- Network Design Impacts: How will end users access data and applications? What network capabilities does the DR service need:
- L2 DC/L2VPN data center extension
- Network Traffic engineering
- Hybrid VPN Services
- Remote access
- Global load balancing
- Pre-configured recovery site networks including IP address mapping, firewall rules & VLAN configuration
- Applications using hardcoded addresses
- Duplicated IP addressing in production and recovery centers
- Security: What are your organization’s requirements for server-level integrity monitoring, encryption for at rest data and network traffic?
- Compliance: Industry specific data retention and data destruction policy requirements.
These strategic considerations will need to be vetted with DRaaS providers as they will affect how to craft a particular DRaaS solution; however, DRaaS providers vary in their capabilities and some only focus on specific industries. Multiple DRaaS offerings have hit the market in the past few years, so it is incumbent on the customer to be thorough in their due diligence to evaluate which DRaaS providers can effectively meet their needs. An IT consulting firm can provide organizations with the necessary insight to prepare DR solution design strategies and assess DRaaS provider solution capabilities. This preparation has proved essential in ensuring DRaaS implementation projects proceed with reduced risks, and result in more cost effective solutions than traditional disaster recovery approaches.
Source: Sky Chat IT Blog