Every day, CEOs ask: How are we better than our rivals? How are we going to stay better? Sound familiar?
For many, the answer increasingly lies in cloud computing. Companies are embracing it enthusiastically, because cloud can help companies be more agile, lower capex, and expand quicker than their rivals. International Data Corporation expects global spending on private cloud services to exceed $24 billion by the end of 2015, making cloud computing practically essential for today’s corporations. While cloud might be widespread, what hasn’t caught on fully is disaster recovery, or DRaaS for cloud.
The cloud alone doesn’t guarantee 100 percent availability of your essential data. A January 2013 vendor survey found that more than 40 percent of surveyed organizations had lost data in the cloud. Fortunately, incidents like this can be avoided through solid back-up and recovery initiatives.
The Year of Cloud Recoverability
This explains why 2015 signals the “Year of Cloud Recoverability.” With more and more data migrating to a virtualized compute environment, companies are depending on their cloud providers to ensure the recoverability of these environments.
More organizations are recognizing their biggest misconceptions about cloud disaster recovery. They are thinking harder about data recoverability, asking more questions of their third-party cloud providers, and getting more involved to ensure the resiliency of their data.
Since the cloud involves a whole network of servers holding your data, its virtualized architecture is geographically dispersed. Therefore, its services are available from many physical data centers. Data resides in one location and is replicated in another or in several other sites with standby servers ready to process that data. That’s sensible since if one site is down and its data damaged or destroyed, the other sites maintain the data. Virtualization provides the tool for recovery. The cloud provider merely ports the data to another location that includes the servers, storage, and network interface.
What could be more secure than that? The problem: it doesn’t always work. Cloud computing can be, well, fragile – and lose things, as the 2013 survey found. While practically all of that data was recovered, it still costs time and some expense to handle the situation. It also makes organizations realize that data that can’t be found can cost lost revenue and reputation and trigger legal issues. Actual data loss in the cloud isn’t infrequent either as the fast clip of data breaches suggest. It’s not just about the data – having data that is safe, without servers to process that data, could still leave companies exposed.
What Organizations Will Ask Themselves and Do
This year and beyond, organizations will be securing the answers to several key questions about resiliency in the cloud and will also take some concrete actions. They will determine how much data they have in the cloud, how often they want to capture it, how long they will need to restore it, and how quickly they will need it.
Their conclusion after that: you can’t treat all gigabytes equally. They will determine that they must set priorities to distinguish what data is most important. In effect, it may be figuring out that one category of data is worth $10 a minute during an outage while another is worth $10,000 a minute. Point-of-sales data may be deemed most vital while HR systems may rank lower on the list. You might put HR data on tape backup while POS inventory might warrant being placed on a virtual machine, or VM, for cloud recovery.
A resiliency workshop with an organization’s applications can help determine just that. So can a business impact analysis. Organizations also will be increasingly determining if, during a system outage, their cloud provider’s services will recover both cloud and legacy systems.
This is important because operations might not return to normal even if all cloud data is recovered. That’s because data stored elsewhere may still be vital to full recoverability. And for organizations that think a cloud service provider such as Google or Amazon can’t go down, think again: they each have had outages. As a result, their customers have suffered. Consider Xbox players who discovered they couldn’t play, for example.
Consequently, more organizations will realize they can’t back up their data in the cloud without considering data recoverability, including full resiliency. Think of it this way: when talking to teenagers, one must impress on them that decisions made today may not show up as mistakes for years to come. An example is if a student decides against taking algebra, only to determine later that if he or she had taken the course, they would be better qualified for a job advancement.
Similarly, organizations also will find that retrieving data from the public cloud especially can be expensive if not properly planned for. Organizations often are captivated by low storage fees in the cloud such as a penny a gigabyte, but they forget that retrieving that data can prove costly, depending on how deep that data is stored.
Retrieving Data Inexpensively
Organizations will increasingly be looking for cloud storage providers who can enable them to retrieve data inexpensively. To do so, it will be important that the provider can dig deep into the application layer when recovering data. Tech-focused companies, in particular, find this service most advantageous: where they can easily get a VM up and access their data. But it is not all about data. Having one’s data safe without a way to get that data to a fully-functional server environment is as good as having a thumb drive without a laptop to put it in. The key is data plus compute equals recovery.
In addition, expect organizations to ask to test cloud recovery services more often to help ensure they can retrieve data when needed. In the past, a cloud service provider might do that twice a year in three-day periods (examples could be June and December). Now, it’s not unusual to see a six-day testing period.
Virtualization: A Buzzword?
Virtualization of an organization’s data has grown leaps and bounds. Throughout 2015, more organizations will move to make sure they can use their disaster recovery and data resiliency site 24/7, for one reason or another.
Let’s say that an organization has an app that runs its business. What happens today is that the IT department will buy a server and make a copy, giving the app site to the developer to write code and see if it breaks. This will become outdated.
More frequently, organizations will spin up their DR site, which becomes their test-and-development location. That’s because their DR service provider will offer a package around doing just that – even making it possible to get multiple copies for use among testers, developers, quality assurance experts, and pre-staging professionals.
This is where the future is heading. The result: greater agility. Organizations will see faster turnaround in getting data copied. For those that introduce new software releases, they will see quicker cycles so they can launch two to three releases a month instead of one.
That will make CEOs happy because if their organizations can spin up on their provider’s cloud and not worry about data resiliency or managing data on physical servers, then they will gain that differentiator they’re seeking. And the CIO can rest easier at night.
Anthony Kessel, senior director, product management, oversees the product lifecycle of Sungard Availability Services’ Recover2Cloud RaaS solution. He has 15 years of experience in the IT product and channel management space. Throughout his career, Kessel has successfully delivered products ranging from traditional professional services, hardware-based sales solutions, and application management services (enabled via traditional, SaaS, cloud, or customized methods). He has worked with both internally developed products as well as partner-delivered solutions with partners including EMC, Cisco, NetApp, Actifio, VMware, IBM, SAP, and HP. Kessel focuses on taking poorly-defined capabilities and creating well-defined products with strong value propositions. Kessel graduated with a bachelor of arts in physics from the University of South Florida. He is ITIL v3 certified.
