Disaster Recovery Journal Winter 2022

2021, when Amazon’s East Coast data center took out dozens of cloud services from chat services to Internet-connected vacuums. Put the right tools in place : It is nec essary to have visibility across all assets, including technology platforms and all other services in use. It is not possible to properly protect your assets if you are not aware of them, and long gone are the days when a spreadsheet was a suitable tracking tool. At a minimum, companies must be aware of their physical systems – e.g., servers, laptops, phones – as well as their cloud infrastructure, services, and data which reside in third-party SaaS services. Automate what you can : Finally, having automation in play and deploying technologies without much human inter vention can help limit how far the disas ter spreads. By removing human error, the management of certain technologies during a disaster event becomes simpler. For example, there are fewer technolo gies that need to be halted or restarted manually; instead, automation does the job of the human hand. The technology system can continue to perform uninter rupted, especially useful in a scenario where employees cannot get to work due to power outages or road closures. Disasters are spontaneous and unpre dictable. As such, companies should pre pare for the inevitable and put disaster recovery plans and processes in place to avoid worst-case scenarios. Plan mes saging must be clear and consistent, and everyone must align on direction, so employees and IT leaders understand their roles and responsibilities when a disaster hits. Streamlining information sharing with all internal stakeholders is critical to solving problems quickly and allows com panies to tackle the disaster head-on, with full confidence. v Jesse Stockall is chief architect at Snow Software. He has more than 20 years of industry experience leading agile teams from concept through to delivery and the adoption of software solutions. He has held previous positions at Symbium, CRYPTOCard, the Canadian government, and Digital Equipment Corp.

68% of IT leaders state they now have a hybrid cloud strategy in place but are experiencing challenges due to their cloud adoption strategy. It’s clear work remains to be done. Adopting a cloud strategy can be chal lenging, even though having the tech nology in place can both better protect companies from disaster and support a work-from-home environment. Care must be taken to ensure an outage at a cloud provider does not have significant nega tive impacts. From data center fires to poorly rolled out upgrades, cloud provid ers have proven fallible. Their disruptions often cause a ripple effect as other services layered on top of the cloud also suffer from outages and a lack of resiliency. How to prepare for the inevitable Disasters happen, that much we cannot change. We can, however, design our sys tems and policies with proper and planned reactions to outages in place to ensure our businesses can continue to operate with minimal disruption. Cloud certainly does not alleviate the need for business conti nuity planning and disaster recovery pro cesses, but it can help make them a bit

smoother and more flexible. The first step is a clear understanding of roles and responsibilities, usually shared between a business and their selected cloud providers. Compared to traditional on-prem models, where a business was responsible for everything in the technol ogy stack, cloud operates with a shared responsibility model. Knowing where the providers’ responsibilities end is crucial to forming a comprehensive strategy and mitigating risk. The shared responsibility model may result in a feeling control is being lost, and this is true to some degree, but it brings new capabilities and opportunities which were not feasible in a wholly on-prem model. Diversify your resources : Cloud providers have data centers in different regions and availability zones. You must understand the specifics of your provider and take advantage of them. A natural disaster may take out multiple data centers close to one another, while a faulty update can quickly cascade across regions. We witnessed how poorly companies lever aged cloud availability zones in December

DISASTER RECOVERY JOURNAL | WINTER 2022 41