tashatuvango - Fotolia

Manage Learn to apply best practices and optimize your operations.

IoT, outages show importance of a cloud backup and recovery strategy

Listen to this podcast

As IoT and big data systems proliferate, IT teams need to store and protect more data in the cloud. To meet those needs, and avoid data loss, rethink cloud backup and recovery.

Cloud backup and recovery have long been a priority for enterprises running production workloads in the cloud. But, today, as trends like the internet of things spur massive amounts of data for organizations to store and protect, IT teams must evolve their cloud backup and recovery strategy -- and make recovery a prime concern.

"Data protection is changing in the world of cloud computing, as IoT comes into play [and] big data systems come into play," says David Linthicum, SVP of Cloud Technology Partners, a cloud consulting firm based in Boston. "We have a lot more data to protect these days."

In this podcast, Linthicum speaks with Tarun Thakur, co-founder and CEO at Datos IO, a data protection software provider based in San Jose, Calif., about the latest cloud backup trends, the disaster recovery lessons learned from the Delta Air Lines outage in August, and why enterprise "cloud denial" is finally fizzling out.

Three trends reshaping cloud backup and recovery

As data from IoT devices and big data systems proliferates in the enterprise, there are a number of ways IT teams should adapt their cloud backup and recovery strategy. But three other trends, particularly in cloud storage, are causing organizations to rethink the way they perform cloud backups.

First, strongly consistent databases, such as an Oracle or MySQL database, are being increasingly displaced by eventually consistent databases, Thakur says. These databases, such as Apache Cassandra, are also distributed systems that scale significantly more than legacy databases.

Second, most next-generation infrastructures, such as those used by cloud providers, use local or distributed storage, rather than shared or file storage. And, third, these next-generation infrastructures are highly elastic, which has a major effect on data protection, according to Thakur.

"If I created a backup of a 9-node cluster, but by the time I have to restore, my cluster size increases to 30 nodes -- how do you manage the policy changes in terms of recovery?" he says. "At the end of the day, it's all about recovery; backup is just a means to an end … and elasticity has a very big impact on how you think about recovery." [5:40-8:35]

What can we learn from the Delta outage?

In August, Delta Air Lines suffered an outage that caused six hours of system downtime, in addition to flight cancellations and delays. The airline's inability to successfully, and swiftly, failover to a backup data center was identified as one of the main triggers behind the disruption -- causing some to question the airline's backup strategy.

"Ultimately, Delta had the responsibility of backing up their system and having an active-active redundancy in place, so if there was a failure of the primary [system], the secondary would be able to take over automatically -- and that didn't happen," Linthicum says.

Other major airlines, including United Airlines and JetBlue, were also affected by data center outages over the course of the past year. For IT pros, these outages present a learning opportunity. Thakur says, namely, don't ever take "shortcuts" when it comes to building a cloud backup and recovery strategy.

While it's natural to want to evolve your IT infrastructure, and experiment with new technologies and services -- including those in the cloud -- "be very careful that, as you're onboarding [these new services] … you do not compromise what is needed to keep your applications and business running all the time, [24/7]," he says. [9:30-16:00]

Cloud denial fades away

While cloud security and compliance concerns still exist in the enterprise, organizations, in general, are much quicker to embrace cloud computing services today than they were even a few years ago, Linthicum says.

"Really, we're getting out of the denial phase and into the operational phase, so people are moving into the cloud no matter what Global 2000 company you work for," he says. "Cloud was there [before], typically as a shadow IT thing, but now it serves as a primary IT initiative, and so this seems to be the trend going forward."

Part of this shift, Thakur says, stems from the fact that cloud is moving beyond just lines of business departments -- that, in some cases, would bypass IT to deploy cloud services -- and becoming a main focus for IT operations teams. As that happens, it's not just the line of business apps that organizations host in the cloud, but their core databases and enterprise applications as well.

While security concerns and other factors still deter some organizations from making the move to cloud, it seems that adoption is not only becoming more common, but more formalized in the enterprise. And that trend, Thakur says, will likely continue.

"There is so much ahead of us," he says. "We are just touching the tip of the iceberg right now." [17:15 – 20:00].

Next Steps

Google cuts cloud storage prices

Using Azure Site Recovery for cloud backup, disaster recovery

Veeam rolls out direct restore service for Azure