Survey Surfaces Significant HA/DR Dissatisfaction Following Outages

A survey of 250 IT leaders in North America and the UK by SIOS Technology Corp. found that 66% experienced at least one outage lasting more than 10 minutes in the past year, yet only 7% test their high availability/disaster recovery (HA/DR) processes monthly. Only 56% are satisfied with their HA/DR platforms, and 75% plan to make DR a medium or high spending priority in the next 12 months.

TL;DR — Key Takeaways HA/DR does not guarantee resilience. Two-thirds of surveyed IT leaders experienced at least one outage lasting more than 10 minutes in the past year. Testing remains a major weakness. Only 7% test HA/DR processes monthly, while one-third test just once a year. DR is becoming more strategically important as infrastructure complexity, ransomware and AI-driven vulnerability discovery increase operational risk. A survey of 250 IT leaders from North America and the United Kingdom finds that despite having a high availability/disaster recovery HA/DR capability in place already, two-thirds 66% had at least one outage lasting more than 10 minutes over the past year, with 23% experiencing three to four such incidents. Conducted by SIOS Technology Corp., a provider of an HA platform, only 56% of respondents said they were either somewhat or very satisfied with their current HA/DR platforms. At the same time, however, only 7% of respondents said their organization tests their HA/DR processes monthly, with one-third testing just annually. Another 22% said they did not know how frequently testing occurs. Dave Bermingham, senior technical evangelist for SIOS, said much of the dissatisfaction with HA/DR solutions stems from that lack of testing. An HA/DR solution that is not regularly tested is just a theory rather than an actual solution. Not surprisingly, three-quarters of respondents 75% label DR as a medium or high spending priority for the next 12 months, with 31% evaluating the role HA might play in patch management as the number of vulnerabilities in applications being discovered by artificial intelligence AI models continues to exponentially increase. Total cost of ownership 35% , reliability and performance 32% and software license price 23% were cited as the top criteria for evaluating a HA/DR solution. The two biggest HA/DR challenges cited were complexity of configuration and management 44% , followed by integration with existing systems and applications 37% . Each IT organization will need to determine to what degree additional investments in HA/DR are required, but as IT environments become more complex, the odds there will be an outage only increase. Additionally, as the volume of cyberattacks continues to increase in the AI era, more IT organizations will view DR as their last line of defense in the event of, for example, a successful ransomware attack. The challenge, of course, is that if the recovery process takes too long, the cost of paying the ransom is often less than the potential revenue an organization might lose. In general, IT teams should be able to recover an IT environment regardless of how much data there is within a day, said Bermingham. Otherwise, something is probably horribly wrong with their approach, he added. Unfortunately, too many IT teams are overly dependent on HA/DR workflows that haven’t been kept up to date. As more applications are added to an IT environment, each subsequent update to that application can have HA/DR implications. As such, there is now a greater need than ever to test HA/DR workflows and processes at least once a month if not more. There should soon come a day, hopefully, when AI tools make HA/DR simpler to implement and maintain. In the meantime, however, IT teams would be well advised to hope for the best while always being prepared for the worst.