Vitamin Y: Cloud storage, backup and recovery

Context:

May’s gathering of the TechConnex cloud peer group was addressed by Dave Clark, business leader for Clear Cover Data Protection, and Matt Adams of virtual data centre provider ThinkOn. InsightaaS recognizes this as a huge growth area within the IT industry: according to survey data from Techaisle, cloud storage is the most widely used cloud service today, and will be for the foreseeable future. Use of cloud for related requirements is also poised for high-growth; we see firms looking to establish cloud-based disaster recovery/business continuity strategies as a way of easing into cloud without disturbing existing investments in on-premise equipment, and as a means of taking advantage of either new capabilities or plunging price points.

In many ways, Clear Cover’s evolution as a business echoes changes in the cloud data management market. The firm started with an automated backup and recovery service based on Asigra, and branched out to offering disaster recovery/data restoration, secure file sharing and application and data hosting as client needs became more varied/sophisticated and as niche profitability became more problematic. Clear Cover’s working relationship with ThinkOn also illustrates the evolution of the market: firms that are successful in delivering cloud have need of relationships with complementary suppliers, a condition highlighted in research conducted last year by Techaisle.

Three keYs:

• The ‘best’ use of cloud storage resources is defined by understanding business requirements. Managers who consider the role of ‘IT IP’ — applications and the data produced/used by those applications — will likely have some common requirements: they will want the process for moving data from one place to another to be as automated as possible; they will want to include all data (from servers, desktops, laptops, virtual systems, cloud systems and smartphones) in that process; they will want to ensure that the process is compliant with existing expectations and regulatory requirements, and additional requirements (e.g., HIPAA) expected to seep into broader use over time. Beyond that, there is the hard but necessary work of identifying the relative importance of different types of information: some data (such as old marketing material) can be safely relegated to low-priority storage systems, while other types (such as ecommerce transactions or payroll data) have more stringent access and security requirements. Different service levels have different infrastructure and management costs; striking an effective balance across data and application categories is important to the overall efficiency of the cloud storage approach.
• The adoption curve starts with questions, not technology. In the presentation, Clark and Adams highlighted a series of questions that buy-side managers should ask and answer before committing to a specific technology or solution approach. These include: are we looking for backup or replication? Do we want the cost advantages of multi-tenancy, or the added control of private cloud? What data should be targeted for local vs. cloud restoration? How will we determine ROI? Once the solution is deployed, will you be able to recover all mission critical applications within the defined RTO (recovery time objective) and RPO (recovery point objective) targets? And very importantly, according to both presenters: what is our recovery testing process? By working through these questions — often, with storage solution experts who can help identify trade-off advantages and disadvantages — buy-side managers can narrow down options to those that meet corporate objectives.
• Restore & recover is a process, not an event. The presenters walked through a multi-step process around restoring and recovering data and applications. They noted that the most common restoration requirement is for message-level content (such as emails). At the next level, critical data is restored at high speed via local storage on a LAN, while cloud archived data can restored via a (generally lower-speed) WAN link. Large/complex structures often require restoration from physical media, with portable disks shipped to the data centre or to an alternative location. The cloud becomes essential at the highest level — where users simply want uinterrupted access to cloud-resident storage systems, or where cloud DR/BC systems provide ongoing operational support in the event of a complete on-premise system failure. As with the previous point, managers who develop clear definitions of the source and priority of restores — who ask and answer questions about business needs associated with IT IP — will be best positioned for success.

Bottom line:

The cloud market for storage, replication, business continuity and related services is complex. It includes a very large number of firms offering a wide range of (often-overlapping) capabilities at many different price points — all of which contributes to confusion on the part of the buyer.

At a fundamental level, though, it’s helpful to remember that there is a distinction between archiving and business continuity. The difference isn’t just one of technical architecture — it is a difference between use cases and anticipated benefits, and which incurs differences in the processes required to support deployment. Archiving is, in many ways, a ‘necessary evil’ — a responsibility imposed by regulators or other stakeholders. As the Figure below shows, this requirement is best met with lowest-cost resources. Business continuity, on the other hand, uses replication to provide for uninterrupted access to technology in the event that an organization’s main servers go down. In some cases — such as in the event of a physical problem with a server or storage — merely providing access to files and applications will be sufficient to ensure continuous operations. In others — for example, a flood or a health warning — access to files and applications is only one part of the continuity strategy; ensuring that there are HR and communications processes in place to deal with physical access issues is also essential to continued operations.

Sitting somewhere between these points is use of cloud storage as an extension of on-premise storage, and/or as a means of exchanging files between colleagues or with suppliers/customers. Again, these use cases can point to different technologies and different deployment and management strategies — and as the lines between these types of services blur, the optimal mix of technology and services becomes harder to define. Buyers are advised to understand the use of IT IP — applications and data — within their operations as a first step, and then to connect technological and management strategies for best protection and availability of the data once the business requirement is clearly understood.

Related reading

Vitamin Y: ASCII Chicago