Is Your DR Plan Actually Setting You Up For Disaster?
By Alex Becker, vice president and general manager of Cloud Solutions, Arcserve
If you’re like most IT professionals, your worst nightmare is waking up to the harsh reality that one of your primary systems or applications has crashed and you’ve experienced data loss. Whether caused by fire, flood, earthquake, cyber attack, programming glitch, hardware failure, human error, whatever – this is generally the moment that panic sets in.
While most IT teams understand unplanned downtime is a question of when, not if, many wouldn’t be able to recover business-critical data in time to avoid a disruption in business. According to new survey research commissioned by Arcserve of 759 global IT decision-makers, half revealed they have less than an hour to recover business-critical data before it starts impacting revenue, yet only a quarter cite being extremely confident in their ability to do so. The obvious question is why.
UNTANGLING THE KNOT OF 21ST CENTURY IT
Navigating modern IT can seem like stumbling through a maze. Infrastructures are rapidly transforming, spreading across different platforms, vendors and locations, but still often include non-x86 platforms to support legacy applications. With these multi-generational IT environments, businesses face increased risk of data loss and extended downtime caused by gaps in the labyrinth of primary and secondary data centers, cloud workloads, operating environments, disaster recovery (DR) plans and colocation facilities.
Yet, despite the complex nature of today’s environments, over half of companies resort to using two or more backup solutions, further adding to the complexity they’re attempting to solve. Never mind delivering on service level agreements (SLAs) or, in many cases, protecting data beyond mission-critical systems and applications.
It seems modern disaster recovery has become more about keeping the lights on than proactively avoiding the impacts of disaster. Because of this, many organizations develop DR plans to recover as quickly as possible during an outage. But, there’s just one problem: when was their most recent backup?
WOULD YOU EAT DAY-OLD SUSHI?
Day-old sushi is your backup. That’s right, if you’ve left your California Roll sitting out all night, chances are it’s the same age as your data if you do daily backups. One will cause a nasty bout of food poisoning and the other a massive loss of business data. Horrified or just extremely nauseated?
You may be thinking this is a bit dramatic, but if your last backup was yesterday, you’re essentially willing to accept more than 24 hours of lost business activity. For most companies, losing transactional information for this length of time would wreak havoc on their business. And, if those backups are corrupted, the ability to recover quickly becomes irrelevant.
While the answer to this challenge may seem obvious (backup more frequently), it’s far from simple. We must remember that in the quest to architect a simple DR plan, many organizations make the one wrong move that becomes their downfall: they use too many solutions, often trying to overcompensate for capabilities offered in one but not the others.
The other, and arguably more alarming reason, is a general lack of understanding about what’s truly viable with any given vendor. While many solutions today can get your organization back online in minutes, the key is minimizing the amount of business activity lost during an unplanned outage. It’s this factor that can easily be overlooked, and one that most solutions cannot deliver.
WHEN A BLIP TURNS BRUTAL
Imagine, for a moment, you have a power failure that brings down your systems and one of two scenarios plays out. In the first, you’re confident you can recover quickly, spinning up your primary application in minutes only to realize the data you’re restoring is hours – or even days old. Your manager is frantic and your sales team is furious as they stand by and watch every order from the past day go missing. In the second scenario, you’re confident you can recover quickly and spin up your primary application in minutes. This time, however, with data that was synced just a few seconds or minutes ago. This is the difference between a blip on the radar of your internal and external customers, and potentially hundreds of thousands (or more) in lost revenue, not to mention damage to you and your organization’s reputation which is right up there with financial loss.
For a variety of reasons ranging from perceived cost and complexity to limited network bandwidth and resistance to change, many shy away from deploying DR solutions that could very well enable them to avoid IT disasters. However, leveraging a solution that can keep your “blip” from turning brutal is easily the best kept secret of a DR strategy that works, and one that simply doesn’t.
ASK THESE 10 QUESTIONS TO MAKE SURE YOUR DR SOLUTION ISN’T TRICKING YOU
Many IT leaders agree that the volume of data lost during downtime (your recovery point objective, or RPO) is equally, if not more important than the time it takes to restore (your recovery time objective, or RTO). The trick is wading through the countless solutions that promise 100 percent uptime, but fall short in supporting stringent RPOs for critical systems and applications. These questions can help you evaluate whether your solution will make the cut or leave you in the cold:
- Does the solution include on-premises (for quick recovery of one or a few systems), remote (for critical systems at remote locations), private cloud you have already invested in, public cloud (Amazon/Azure) and purpose-built vendor cloud options? Your needs may vary and the solution should offer broad options to fit your infrastructure and business requirements.
- How many vendors would be involved in your end-to-end DR solution, including software, hardware, networking, cloud services, DR hypervisors and high availability? How many user interfaces would that entail? The patchwork-based solution from numerous vendors may increase complexity, time to manage and internal costs – and more importantly it will increase risks of bouncing between vendors if something goes wrong.
- Does the solution provide support and recovery for all generations of IT platforms, including non-x86, x86, physical, virtual and cloud instances running Windows and/or Linux?
- Does the solution offer both direct-to-cloud and hybrid cloud options? This ensures you can address any business requirement and truly safeguard your IT transformation.
- Does the solution deliver sub five-minute, rapid push-button failover? This allows you to continue accessing business-critical applications during a downtime event, as well as power on / run your environment with the click of a button.
- Does it support both rapid failover (RTOs) and RPOs of minutes, regardless of network complexity? When interruption happens, it’s vital that you can access business-critical applications with minimal disruption and effectively protect these systems by supporting RPOs of minutes.
- Does the solution provide automated incremental failback to bring back all applications and databases in their most current state to your on-premises environment?
- Does your solution leverage image-based technology to ensure no important data or configuration is left behind?
- Is your solution optimized for low bandwidth locations, being capable of moving large volumes of data to and from the cloud without draining bandwidth?
- In the event of a disaster, does the solution give you options for network connectivity, such as point to site VPN, site to site VPN and site to site VPN with IP takeover?
The true value you provide your organization and your customers is the peace of mind and viability of their business when a disaster or downtime event occurs. And even when its business as usual, you’ll be able to support a range of needs – such as migrating workloads to a public or private cloud, advanced hypervisor protection, and support of sub-minute RTOs and RPOs – across every IT platform, from UNIX and x86 to public and private clouds.
By keeping these questions in mind, you’ll be better prepared to challenge vendor promises that often cannot be delivered and to select the right solution to safeguard your entire IT infrastructure – when disaster strikes and when it doesn’t. No more day old sushi. No more secrets.
About the Author
As VP and GM of Arcserve Cloud Solutions, Alex Becker leads the company’s cloud and north american sales teams. Before joining Arcserve in April 2018, Alex served in various sales and leadership positions at ClickSoftware, Digital River, Fujitsu Consulting, and PTC.
Source: Disaster Recovery Journal Is Your DR Plan Actually Setting You Up For Disaster?