Video: 10 Minute IT Jams - StarWind engineer on the optimal disaster recovery approach

Disaster recovery is not just a buzzword. It's an urgent necessity for modern businesses navigating an unpredictable digital landscape. In a recent interview on 10 Minute IQ Jams, StarWind Solutions Engineer Orris Lester explained why proper strategies for data protection are more critical than ever – and how companies can guard against threats like ransomware.

Data resilience is often misunderstood, Lester warned, with many confusing key concepts. "These are actually the terms that often get mixed and misinterpreted for various reasons," he said, addressing the differences between RAID, backup, and disaster recovery.

RAID, or Redundant Array of Independent Disks, is a technology designed to "collect the drives into a single or multiple logical units, distribute data over the drives, and add redundancy depending on the RAID level set," Lester explained. The intention is to keep systems running smoothly and avoid downtime if a drive fails. However, he cautioned, "It's not the separate data copy. The main idea behind RAID is to ensure uptime and performance for the system. So it protects the system against the drive or drives failure. But when it comes to ransomware or even simply if you delete a file by mistake, your data is gone."

By contrast, a backup is a distinct, time-specific copy of data, stored for as long as needed. "Backup allows you to restore and recover your VMs, your data, whenever it's needed or in the worst case scenarios," Lester noted. However, he clarified that backup is not primarily about the fastest possible recovery: "It's more about the amount of data you can access whenever needed. For example, you can store the backups for years."

Disaster recovery (DR), according to Lester, is about "exactly the shortest recovery time objective and recovery point objective." This usually entails keeping a secondary copy of data at a remote location, with regular asynchronous replication in case the primary site goes down. DR is designed for fast, near-seamless resumption of services in a catastrophe.

Most businesses, Lester said, find it difficult to implement both systems independently due to costs and complexity, but this isn't necessarily problematic. "All depends on the company's requirements and RTO and RPO," he added, referring to Recovery Time and Recovery Point Objectives.

So, what should customers prioritise when choosing solutions? "The first thing I would point is the ability for the backup solution to easily integrate within the infrastructure without you having to reinvent your backup plan, your backup schedules and backup windows," said Lester. Integration, he argued, should not create a bottleneck nor impede production workloads.

Quick recovery is also vital: "You'd want the backup solution to at least be able and be ready to restore fast, basically as fast as possible in the worst case scenarios or if needed," Lester emphasised.

A cloud component is increasingly crucial as well, enabling an "additional copy of your data in the cloud, which can very well serve instead of the remote off-site facility – and again increase your resilience against ransomware."

Ransomware itself remains a looming concern in disaster recovery planning. Lester believes there's "no single piece of software or hardware to ensure 100 percent protection against ransomware or 100 percent guarantee you'll get your data recovered." Instead, he advocates for a combination of practices.

According to Lester, companies should "use immutable storage whenever possible, be it on premises or in cloud. It will at least provide you the ability to keep the backups intact and restore from them, of course if the backups were performed prior to the ransomware attack." He also encouraged running "weekly, monthly, even maybe quarterly and yearly separate backup jobs and keeping them separately. Yes, it requires more storage, it's more expensive correspondingly, but again it's about increasing your chances to restore a healthy and not encrypted VM from one of these backup copies you already have."

But technology alone isn't enough. Training remains crucial. "It's the actual training of your staff and employees, because it's about the processes. They should be aware of what steps they need to take when the ransomware hits and even most importantly to be informed of the potential sources of the ransomware," Lester said. He observed that, more often than not, successful attacks come from simple employee mistakes rather than highly targeted assaults. "That's the role of the preparedness and of the proper training for the staff – to be always aware and how they should act to minimise again the risks of the ransomware commission."

Naturally, the interview turned to how StarWind's own product fits into this ecosystem. "We designed our backup appliance with the goal of basically taking traditional backup infrastructure to an entirely new level," Lester stated. The company's backup appliance is engineered to avoid interfering with production environments: "It's basically a level where you don't have to thoroughly think of and tinker with the backup windows and schedules, and without having to worry about your backup process overlapping with the production hours."

A key differentiator for StarWind's offering is its use of all-NVMe storage for both speed and resilience. "StarWind backup appliance is the only backup appliance on the market that uses all NVMe storage." NVMe, Lester said, is "extremely fast", reducing backup windows and enabling quicker recoveries. Another innovation is that the backup target "connects the backup target when the backup job is started and disconnects it when it's finished, so all the rest of the time when the backup jobs are not running, the backup target is basically offline, which minimises the risk of ransomware getting to the storage and encrypting your backups."

For organisations with strict recovery time and recovery point objectives, this can underpin more robust data protection. "This becomes the perfect solution for the companies that have strict recovery time and recovery point objectives, maybe allowing them even to perform more backups during the time when they previously managed to perform only a single one, thus giving them ability to ensure better RPO," said Lester.

He also pointed to proactive support as a key feature: "It monitors the appliance itself for the hardware and software components and even more important for the backup jobs to make sure there are no, literally no to minimum possible, issues with backups," he said.

In the end, a layered mix of technology, process, and people keeps data safe in the face of growing threats. As Lester concluded, "Thank you very much."