Is RaaS the next frontier for DR?

BANGALORE, INDIA: Disaster Recovery (DR) is a space that has witnessed new contours every few years. As technologies come and disrupt the landscape and as economic forces inject new imperatives for businesses; the nature, scope and model of an apt DR plan keeps taking new shapes.

The latest inflection point to watch in this graph has been an irony-filled dilemma - of addressing unrelenting uptime needs of businesses in today's staggering pace along with expectations of doing so without piling up capex redundancies and unreasonable budgets.

"Most companies have to focus on business and that's why they have to turn to other people who can provide the expertise and the solution desired, letting businesses, in turn to focus on their core. If someone can do that with an approach which leaves no room for security worries and other issues, that's a strong answer," in the words of Keith Tilley, EVP EMEA & APAC, SunGard Availability Services.

Now when he says that ‘back-up is not recovery' as he explains a new-era proposition, it leaves us a tad curious. "You need a holistic approach and back-up is just one slice of that. Plus, why to pay too much when a comprehensive plan is possible?" he quips.

The company, in India, is in various stages of engagement with multiple prospects with the RaaS offering, while in mature markets it has in its kitty hundreds of customers who are utilizing RaaS offering, and as Tilley confidently puts it, the list continues to grow.

As we chat with Bala Vasireddi, SVP & GM of India Operations at SunGard Availability Services, let's find out more about the concept as well as its potential while we also understand changing tides around consumerisation-propped DR, Tape storage, WAN, Virtualisation and other areas that flank DR.

DR as a Service - now why does that sound like a paradox? Is it possibly because we have always seen DR synonymous with an extra server, data center or location back-up?

The unique thing about DR is that you won't know its value unless an incident happens. India may not have seen many exigencies but being prepared is a must. Small disasters happen every day and ability to survive data loss is critical. At SunGard Availability Services we have come up with a unique offering on Recovery as a Service and it stands out for all the contractual guarantee terms, SLAs etc that we care to weave in unlike other players. We are addressing a problem that has already popped up in mature markets earlier. When you look at customers who have been in this area with critical IT, we also see how IT environments have become complex especially for verticals like BFSI, Telco, healthcare etc.

So far there was a lack of service providers. There was need for someone who can put across a fully-reliable, resilient plan that helps them do away with capes as cost pressures increase. People need to understand the level of complexity and resilience that is a key and hard-to-ignore factor. What is available out there or possible internally can be adequate, but customers need to ask if it is fast enough, cheap enough and a sure-shot answer with the kind of RPO (Recovery Point Objectives) needed?

Elaborate.

Most companies say they have done testing but it is usually limited in scope and scale, and does not factor in other dependencies. When you do something as part of production, it is not a full-fledged testing. We see this problem in international markets also. Production people have limitations when it comes to DR especially with the kind of uncertainty and risks involved. Now who can a CIO hold accountable? That's where one needs techniques in place to go into customer environments (and by the way, every environment is different) to understand them, put mechanisms in place for guaranteed recovery metrics which we are contractually committed to provide. Our team of global experts also brings in a best-practice flavor which is not possible with conventional DR models. It is not about just IT skills here but about best practices across the market.

 

How do metrics like RTOs (Recovery Time Objectives) and RPOs play out in this scenario?

You can think of RTOs around four hours or less and almost zero RPOs, but depending on environment. Our environment can support multiple centers around the world as more or less there are certain IT problems that are common across the pack. When you are thinking of a business need like crucial RPO, internal back-up is not enough as it does not allow an enterprise to do things cheaper, faster and with agility. Usually production support team and DR staff are at different levels. It's like accounting - while one person can be great in maintaining them, there might be need for another better person when it comes to auditing them. Once-every-six-months- now in my opinion, that's nowhere close to resiliency and yet businesses are satisfied with that.

Is data replication a strong element here? What implications have WAN optimization and Virtualization brought on the DR landscape?

Yes, that is an integral part. WAN optimization improves network throughout and in turn positively impacts the effectiveness of DR that uses Recovery as a Service (RaaS) or for that matter, any other means that relies on the company's network. RaaS, for instance, replicates data in real-time from customer site into a recovery environment, and therefore any optimization gained in the transport layer will improve efficiencies.

Virtualization's inherent strengths help improve service effectiveness in DR. At the time of Disaster (ATOD), virtualized environments can be restored fairly quickly as opposed to physical ones and that; in effect will improve Recovery Time Objective (RTO) of the service. Moreover, Virtualization does enables us to deliver RaaS more cost-effectively due to its inherent ease of configurability, like, share-everything, share-something and share-nothing models. RaaS offering has been designed to work with both, standalone physical servers as well as with any of the virtualized models. Some customers opt for dedicated physical services due to security/privacy concerns while some other choose"flexible" virtualized models for cost benefits, without compromising on security and privacy. Notably, today some business' insistence on a physical server dedicated to their DR is an issue of comfort factor rather than tangible differences in security between virtualized and conventional computing.

So what's your take on the future of tape-as-a-recovery format?

Tape as a recovery format is expected to continue, but only for data that is not mission-critical. While there are several variables that impact recovery time, a 24 - 48 hour timeframe can be taken as a thumb rule. Since this is not acceptable in today's businesses, tape cannot fulfill DR objectives for critical data and applications. Our recommendation is to plan restore/recover for top tier mission critical systems with real time DR offering, thus minimizing data loss and downtime in disaster situations.

However, for not-so-critical data and applications with accepted higher recovery times, tape backup and recovery is an inexpensive yet effective way of facilitating disaster recovery and will continue to do so in near future.

How significant is this model for today's market and your offerings?

Enterprises now have the option of purchasing Disaster Recovery as a service as opposed to investing in it as a product. Consequently, specialists with expertise in effectively analyzing and managing mission critical functions and business continuity needs of an enterprise as a service are taking center stage. Our service offering supports all major industry applications and custom applications across all major platforms and storage / backup technologies. With the objective of replicating its global success in offering DRaaS to enterprises in markets around the world, SunGard Availability Services is introducing its Disaster Recovery as-a-Service (DRaaS) as a new line of business in India targeted at the large enterprises segment.

How does the concept work in juxtaposition to internal DR IT or the models currently available in the market? What about compliance facets?

 

The DIY (Do It Yourself) guy needs confidence to recover data when a real disaster strikes, and that can be inadequate at times, more so when there is lack of full-fledged testing. For others, System Integrators have been there in the market but then ownership is a valid issue and so are discipline and reliability and assurance of guarantees. We support those needs with services tested on a global instance and tested well in India with our strong centre here. Compliance is a key driver apart from audits or need for best practices. For our side, we observe all major regulations and specific mandates for each country, putting global best practices and scenarios across the spectrum, except for minor exceptions for individual location differences.

Are your concerned with the BYOD onslaught happening in DR area with the likes of Google Drive, Dropbox etc?

They are interesting solutions but when you pick them there is a long T&C (Terms & Conditions) footnote which clearly spells out the reality on guaranteed recovery. For us, on the other hand, SLA-based approach is key to allow customers to understand the terms. Further, data recovery is not application recovery. If someone is ok with being clueless as to where their data is stored, in which country, then consumer versions of DR are a different breed altogether.

Think sharply as an enterprise and the scenario is very different. If you are okay for BCP (Business Continuity Planning) to be just a tick in the box, then that's another choice. Till the time a business hits an incident, it may not take it very seriously. Businesses have to learn that DR is not a burden but the cost of doing business today when even the slightest downtime can have far-reaching ramifications. DR is not a cost but a weapon.