DRJ Fall 2019

Conference & Exhibit

Attend The #1 BC/DR Event!

Summer Journal

Volume 32, Issue 2

Full Contents Now Available!

Tuesday, 07 May 2019 18:10

Why Artificial Intelligence-Powered Backup is the Future of Backup and Disaster Recovery

Written by  DON FOSTER

Foster1Over the past few years, businesses have increasingly turned to cloud, mobile, IoT, artificial intelligence (AI), machine learning (ML) and other new technologies to digitally transform the way they create and deliver value. This digital transformation has made it essential for enterprises to be able to reliably access and activate ever-increasing volumes of data if they hope to realize their business objectives. At the same time, enterprises are increasingly seeking to reduce the amount of IT infrastructure they use to process, store and manage data, complementing or even replacing their on-premises datacenters with Software as a Service (SaaS) applications and cloud services. Finally, with governments and customers demanding that companies secure personal data, the risk involved in failing to protect data from increasingly sophisticated ransomware and other cyber threats continues to rise.

These three trends – growing volumes of data that are now essential to an enterprise’s success, more diverse, complex and distributed IT infrastructure, and a riskier threat landscape – have made data backup and recovery a critical strategic business function. However, IT professionals are recognizing that traditional methods of backing up and recovering data are no longer able to adequately perform this function because these technologies were not designed to handle the complexity of today’s IT environments. Enterprises are finding these technologies make quick and complete recovery of critical data from a wide variety of on-premises and cloud-based infrastructure after a cyberattack or other disaster increasingly time-intensive and complex. Unless IT professionals find a way to simplify data protection, the difficulties they are now experiencing achieving the faster and more comprehensive data recovery objectives required in today’s digital economy will only continue to multiply. They will also be caught flat-footed when innovative new technologies emerge, unable to use these technologies to stay competitive in tomorrow’s economy. They will be forced to continuously increase the amount of time, people, and other resources they invest in managing traditional backup and recovery technologies.

Fortunately, some of the same technologies that are driving the need for faster, more comprehensive data recovery readiness also enable enterprises to implement a new type of backup that can achieve this readiness without breaking the bank: self-driving backup. Self-driving backup uses AI and ML technologies to automate backup and recovery operations and management, including setup, monitoring, and service level agreement (SLA) tracking. In doing so, self-driving backup empowers IT professionals to set the data protection outcomes they require while also continuously optimizing backup and recovery operations to realize these outcomes.

By making data protection outcome-based, automating basic backup and recovery care and feeding tasks, and only notifying IT professionals when their intervention is required to keep SLAs on track, self-driving backup provides IT professionals with the agility they need to focus more on higher level data protection, management and governance tasks. In addition, by helping them optimize backup and recovery, IT professionals can recover data faster and more comprehensively, increasing their customers’ trust in their companies’ digital services without dramatically increasing costs.

AI and ML – The Smarts Behind Self-Driving Backup

How does self-driving backup use AI and ML to make data protection outcome-based and automate backup and recovery operations, including queuing, load balancing, scheduling, prioritizing, capacity planning, and storage optimization? It uses AI and ML’s ability to understand data patterns to identify when backup and recovery processes are trending toward failure to meet objectives, then adjusts and updates these processes and operations when it identifies these abnormalities. This capability does require some “training” of the AI and ML technologies before it is fully effective. First IT professionals must install the self-driving backup system and establish their SLAs and other data protection goals. The system then begins completing the tasks necessarily to realize these goals, using its built-in automation and processes. However, as it does so it starts to gather and analyze information on these tasks, how well they are preforming, and if any SLAs are being missed (or might be missed in the future) outside of the user’s defined windows. Using this information, self-driving backup can reallocate resources or otherwise change the way it completes its tasks, resulting in fewer notifications to IT professionals that SLAs or other goals are at risk of not being met. In addition, over time the self-driving backup system gains an understanding of the enterprise’s overall IT environment, its resources and the performance of its resources. When the IT professional wants to add new workloads or systems, it can predict if and how it can achieve various SLAs for these workloads and systems. This understanding is not static – as a self-driving backup system gathers more information on the enterprise’s data protection operations, it can uncover new patterns and better identify anomalies, becoming “smarter.”

For example, after a few weeks of using a self-driving backup system, it can analyze current backup behavior and patterns to predict, with a high degree of accuracy, when the enterprise’s storage will run out and new capacity will be required. IT professionals can use this information to make high-level decisions regarding their enterprise’s data protection strategy to address this prediction – change the amount of data they are backing up, how long they are retaining different types of data, move some backups to the cloud or other infrastructure, or add cloud or on-premises backup capacity. What these professionals are not focusing on is tedious daily backup and recovery tasks. The self-driving backup system executes these in the background, optimizing them to reflect the enterprise’s environment and only alerting users when there is a problem.

Self-driving backup can provide similar insights regarding SLAs and other IT professional-established backup and recovery outcomes. For example, if the system does not meet a specific SLA within a given window, it can automatically update its backup processes to meet the SLA. If it determines that such "tuning" of backup processes will still fail to meet the SLA goal, it will then alert users that they need to evaluate the workload while also providing analysis on why the system is not able to achieve the goal (the amount of data being backed up has grown, more throughput is required, etc.). This process also works in reverse. If a workload’s SLA goals are scaled back, a self-driving backup system can reallocate resources previously needed to meet the SLA for that workload to other operations, improving overall backup and recovery performance.

A Fundamental Shift In How IT Professionals Protect Data

While enterprises once needed to regularly bring in consultants to provide analysis on how to fine-tune backup and recovery operations to be more efficient and meet SLAs, self-driving backup systems automatically deliver these insights on a continuous basis. In addition, self-driving backup streamlines error reporting. Though all errors are logged, the system is intelligent enough to determine which errors are the ones impacting performance and only alert users to those errors. At the same time, self-driving backup can alert users to errors or other irregular activity that, while not currently impacting their SLAs, might indicate a cyberattack or a performance problem in the future. Such anomalies can include low-priority workloads taking much longer than usual to complete, an unusual number of file changes, deletions, moves in a given time period, or honeypot file changes that indicates a potential cybersecurity attack.

These capabilities represent a fundamental shift in how IT professionals protect data – from initiating and carefully monitoring everyday backup and recovery operations, to supervising the self-driving backup system that completes these operations and assists it when it encounters a problem, using insights generated by the system itself to solve the problem. The result is backup and recovery that manages to be both simpler and more sophisticated, delivering IT professionals the time, visibility, and knowledge they need to spend more time on the high-level data recovery challenges their enterprises face as they strive to be leaders in today’s digital economy.

Backup and Recovery That Works For You

Self-driving backup empowers IT professionals to step off the relentless traditional data and recovery solution treadmill and stop constantly investing more money and time in an effort to maintain SLA levels as data volumes grow, applications proliferate, and infrastructure diversifies. Instead, with AI and ML-powered self-driving backup, they have a system that can automate the daily backup and recovery operations they need to meet their SLAs and other data protection goals. They have a system able to learn over time how to optimize and adjust backup and recovery resource allocations to best meet their data protection goals while dynamically updating operations as their workloads and infrastructure evolve. They have a system that alerts them when an SLA might be at risk, or an error or anomaly requires their attention.

For years, cloud, AI, ML, and other new technologies have helped simplify infrastructure scalability, application development, data analysis, and other aspects of IT while complicating data protection. With AI and ML-powered self-driving backup, IT professionals are able to use these technologies to simplify data protection with a system that, rather than requiring more work, actually works for them.

DonFosterDon Foster is senior director of worldwide solutions for Commvault. Foster is a 20-year veteran in the storage, infrastructure, cloud, and data management space. He has been with Commvault for 16 years, holding a variety of technical management roles throughout that time.