Editor’s note: This article, originally published on March 27, 2014, has been updated to more accurately reflect recent trends.
Wildfires in California. A snowstorm in Texas. Windstorms across the Midwest. Floods in Hawaii. Hurricanes in Florida and Louisiana. Russian hackers and ransomware attacks. And let’s not forget the global pandemic.
If anyone still thinks that having a disaster recovery and business continuity plan isn’t a high priority, you haven’t been paying attention to recent events. As we begin to emerge from the COVID-19 pandemic, organizations are shifting to a new normal that will certainly be more remote, more digital and more cloud-based. Disaster recovery plans will have to evolve to keep up with these changing business conditions.
On top of that, business requirements for disaster recovery have changed dramatically. There was a time when it was acceptable for recovery time to be measured in days or hours. Now it’s minutes. In some cases, business units are demanding zero down time in the event of an unplanned outage.
Here are the basics of a state-of-the-art disaster recovery/business continuity (DR/BC) plan for 2021 and beyond. (Without getting too hung up on definitions, let’s say that disaster recovery is getting the IT infrastructure back up and running, while business continuity is a broader discipline that gets the business back up and functioning once the lights are back on.)
For CISOs, the first goal of a disaster recovery plan is to avoid the disaster in the first place, which is becoming increasingly challenging. First, data is no longer safely tucked away in an on-premises data center. It’s distributed across on-premises environments, hyperscale clouds, the edge and SaaS applications. ESG Research Senior Analyst Christophe Bertrand points out that SaaS presents a serious data protection and recovery challenge because “now you have mission critical applications running as a service that you have no control over.”
Second, the pandemic drove millions of employees out of the secure confines of the corporate office to their home offices, where the Wi-Fi is less secure and where employees might be sharing sensitive data on collaboration applications.
Third, hackers took notice of these expanding attack vectors and launched a barrage of new and more targeted ransomware attacks. According to the Sophos State of Ransomware 2020 Report, hackers have moved from spray-and-pray desktop attacks to server-based attacks. “These are highly targeted, sophisticated attacks that take more effort to deploy. However, they are typically far more deadly due to the higher value of assets encrypted and can cripple organizations with multi-million dollar ransom requests,” according to the report.
In response to these changing conditions, CISOs should focus on beefing up endpoint security for remote workers, deploying VPNs and encryption, protecting data at rest no matter where it lives, and also making sure that collaboration tools don’t become a source of security vulnerabilities.
Organizations need to conduct a thorough business impact analysis to identify and evaluate potential effects of disasters through the lenses of financial fallout, regulatory compliance, legal liability, and employee safety. Gartner estimates that 70% of organizations are making disaster recovery decisions without any business-aligned data points or based on an outdated BIA. “Without the fact base the BIA provides, teams can only guess at the appropriate level of DR and what risks are tolerable. This results in overspend or unmet expectations,” according to Gartner.
Remember, you don’t need to protect everything. Organizations that conduct these exercises are often surprised to discover servers that do nothing but run a routine back-end business process once a month, or even once a year.
Organizations need to prioritize applications by their criticality to the business, and to identify all the dependencies associated with a business process, particularly applications that may have been virtualized across multiple physical servers, might be running in containers in the cloud, or in serverless cloud environments.
Along the same lines, you don’t need to protect all data, just the data that you need to keep the business running. You do need to go through the process of locating, identifying, and classifying data. Be sure to protect data that falls under regulatory requirements, customer data, patient data, credit card data, intellectual property, private communications, etc. The good news is that tools can automate data identification and classification.
DRaaS is an increasingly popular option for CISOs at small- to mid-sized organizations who want to cost-effectively improve IT resilience, meet compliance or regulatory requirements, and address resource deficiencies. The DRaaS market is expected to grow at a rate of 12% a year over the next five years, according to Mordor Intelligence. DRaaS services cover the full gamut of disaster recovery and business continuity, providing flexibility and agility to enterprises, according to the Mordor report.
Gartner adds that as the DRaaS market has matured and vendor offerings have become more industrialized, the size and scope of DRaaS implementations have increased significantly, compared with a few years ago.
Simply getting servers back up and running is essentially meaningless unless everyone knows their roles and responsibilities. Do people have the appropriate cell phone numbers and email addresses to share information? Do the relevant stakeholders have a playbook that spells out how to respond to a crisis in terms of contacting law enforcement, outside legal teams, utility companies, key technology and supply chain partners, senior leadership, the broader employee base, external PR teams, etc.?
Depending on the nature of the disaster, networking groups might need to establish new lines of connectivity for remote workers and reconfigure traffic flows; maintenance teams might need to perform remote troubleshooting, security teams might need to re-set firewalls, change access policies, extend security protection to new devices or to cloud-based resources. The biggest problem in a disaster isn’t related to data backups, it’s not having the right people in place and understanding all the steps required for the business to recover, says Bertrand.
To test disaster preparedness, companies traditionally conduct tabletop exercises in which key players physically come together to play out DR scenarios. However, only one-third of organizations perceive the exercises as “highly effective,” according to a July study by Osterman Research in association with Immersive Labs, a company that develops human-readiness skills in cybersecurity. The research also found that organizations don’t perform tabletop exercises often enough to keep up with evolving threats and that these exercises cost an average of $30,000. During the pandemic, it’s fair to assume that tabletop exercises fell by the wayside.
