Download the authoritative guide: Enterprise Data Storage 2018: Optimizing Your Storage Infrastructure
Data protection has suddenly become hot once again. After years languishing as an unexciting item on every storage managers to-do list, it has once again moved to top of mind. This may be a sign of the times: growing cloud adoption, the rise of the distributed enterprise and the demand for analytics are all driving data protection to the forefront. Here are some of the top trends in this area.
Bob Hammer, CEO of Commvault, believes the market is undergoing a fundamental shift. The traditional role of IT was to put up the infrastructure, then figure out how to fit in the apps and the data. But that may not work anymore.
“We are seeing a shift from an infrastructure-centric to a data-centric view,” said Hammer. “In the past, data was centralized and now it is distributed.”
Moving Away from Infrastructure
As a consequence, the old priorities are giving way to the new. For decades, the focus of IT has been on storage, compute power and networking. Those things remain important. But they will are losing ground in importance to data.https://o1.qnsr.com/log/p.gif?;n=203;c=204650394;s=9477;x=7936;f=201801171506010;u=j;z=TIMESTAMP;a=20392931;e=i
Instead of IT, we are seeing the emergence of data technology (DT) and that begins with it data requirements,” said Hammer. “Data management is about sharing dispersed data.”
With data moving front and center, cyber security also moves to the forefront. Cyberattacks are on the increase. Ransomware infiltrates systems. Viruses can slip inside easily and cause havoc. The growing risk of a cybersecurity breach is something that no organization can afford to ignore.
“Organizations have to carefully manage cyber risk,” said Hammer. “Cyberattacks are targeting your data, not your infrastructure.”
Compliance Needs Grow
Storage professionals have been grabbling with compliance requirements for some time. SOx, HIPAA and other regulations have caused them many sleepless nights. They have had to figure out data retention, archiving, and storage architectures that could satisfy these regulations. And it’s only going to get worse. The European Union’s (EU) General Data Protection Regulation (GDPR) has greatly strengthened and unified data protection for all of the EU.
“GDPR has serious repercussions. Anyone doing business in EU is subject to a 4% fine of their revenue if they violate it,” said Hammer. “Most companies are in big trouble as they only have access to 50% of their data, so could fall afoul of this new regulation.”
Analysis at the Heart
Perhaps the most difficult change that storage managers will have to accept is that storing data is no longer enough. Organizations want to mine that data for insight into how they can operate more efficiently. Analytics, then, is rapidly becoming the heart of organizations. They want enhanced search, better content management and the ability to dive into that data to achieve better decision making.
“Analytics is about getting the most use and value from your stored data,” said Archana Venkatraman, an analyst at IDC.
Will Hayes, CEO or Lucidworks, concurs. The key to successful data implementation, he said, isn’t just to provide access, and arrange data in nice-looking dashboards. It is about insight, and how to analyze rapidly as the volume of data growth accelerates. The quantity of data, then, means traditional storage systems need help. Machine learning is being looked to as the answer to this problem.
“The key to leveraging data is machine learning,” said Hayes.
Big Data Failures
The advent of big data analytics may be exciting. And many are embarking on that journey. But it’s far from certain that they will successfully arrive at their destination. Some big data projects fail due to lack of management and policy. But many don’t even get that far.
“Most big data projects fail at load,” said Hayes. “They set up the system and then they can’t get data in.”
Hayes advice is to prepare and manage data better. That requires auto-tagging, classifying and enriching it so you can find it and add intelligence. Once accomplished, the next step is to establish workflows to trigger alerts and unearth insights.
“The last step is remediation,” said Hayes.
This is where innovation is taking place in storage. More and more storage vendors are finding ways to introduce analytics.
One of the big criticisms of cloud storage in the early days was the lack of security. To entice enterprise customers, therefore, cloud vendors have gone to great lengths to address this thorny subject. They followed that up with a PR campaign to promote how well they protect data. But assurances about the security of the cloud may have caused some to become complacent. Many companies threw their storage into public clouds without taking the time to add their own layers of protection.
“It’s a myth that your data is protected in the public cloud,” said Karl Rautenstrauch, Senior Program Manager for Azure Storage at Microsoft. “It isn’t protected, archived., migrated, indexed or searchable. There are no magic fairies that protect public clouds.”
Azure stores over 800 PB in the cloud, with 50 PB more being added each month. Rautenstrauch’s advice is to It backs up your own data (Azure uses Commvault to protect internal data), and carry out data reduction via deduplication and compression.
Backup Becomes Recovery
Jason Buffington, an analyst at Enterprise Strategy Group, said people are less concerned with backup for its own sake, these days. They are interested in recovery as an outcome rather than backup as a method, he said.
“Backup morphed to data protection,” said Buffington. “To backup has been added replication and snapshots to increase recoverability agility.”
Primary Data More Important
Data protection used to focus on secondary data. You did backups then tried to look after them: store them, send them offsite and so on. But the result of that approach, said Huffington, is you end up with lots of copies.
“Start data management with the primary copy, not the secondary copies as in backup,” said Buffington. “If you do it the other way, you end up with an average of 11.3 copies of data for protection per organization, according to our research.”
Buffington is also a fan of data retention policies. He said more and more companies are only holding onto data as long as they have to according to legal or regulatory responsibilities – and no longer.
“It’s best to get rid of what you don’t need via archiving and deletion of data,” he said. “If data is stagnant, don’t waste time recovering it during a crisis.”