Why Application Resilience Fails Despite HA and DR | Natalia Casey, SIOS Technology | TFiR

Swapnil Bhartiya: Despite all these HA and DR solutions, practices in place, mission-critical applications keep going down at organizations that already invested in high availability and disaster recovery. A new SIOS survey found that 76% of organizations had at least one outage lasting more than 10 minutes in the past year despite having HA and DR protections in place. So, The protection exists, but whether it works or not is not clear. What is clear is that the downtime is real. So where does application resilience actually break and what can IT teams do about it? And to discuss that, we have with us once again Natalia Casey, Director of Marketing at SIOS Technology. Natalia, it’s great to have you on the show.

Natalia Casey: Thank you, Swapnil. It’s great to be here.

Swapnil Bhartiya: Your 2026 State of Application Resilience survey found that 76% of organizations had at least one outage that lasted more than 10 minutes in the past year, even when they have HA and DR protection in place. What does that tell you about the state of application resilience today?

Natalia Casey: Yeah, it shows that having the tools isn’t enough anymore. Organizations are running their mission-critical applications in complex hybrid or multicloud environments, and these legacy tools just can’t keep up with these modern workflows. So they can’t keep up with the complexity, and then we’re not doing enough testing. So the systems are complex, the testing isn’t— is infrequent. So that’s why organizations are still experiencing unexpected downtime. The tools aren’t right and the testing isn’t enough.

Swapnil Bhartiya: Nearly 7 in 10 organizations in the survey run critical applications across hybrid environments. How is that changing the way IT teams need to think or rethink about high availability and disaster recovery?

Natalia Casey: Yeah, today’s IT teams aren’t managing isolated servers. They’re managing and keeping applications running across full hybrid environments. So when the systems are that spread out, you need a single unified tool that automates failover across every platform regardless of the cloud or operating system. So they want tools that can that are flexible enough to work with anything in your environment.

Swapnil Bhartiya: One finding that stood out to me was complexity. It has overtaken cost as the biggest HA and DR challenge. Why is managing resilience becoming so complicated?

Natalia Casey: Yeah, managing resilience is complicated because modern IT environments are. Teams are juggling hybrid setups, multiple clouds, and mixed operating systems all at once, all in one environment. So connecting all of those moving parts is complicated. It can be complex. It could be a complex issue to solve. So I think that’s why it has created more of a hurdle than cost or budget.

Swapnil Bhartiya: Survey also says that only 7% of organizations test their disaster recovery plans monthly, and 22% don’t even know how often testing happens. What would it take to make DR testing routine and not a rarity?

Natalia Casey: Yeah, regular testing is non-negotiable. Without it, a disaster recovery plan is just a theory, and it is a living, breathing document. So as your organization adds new technology to it, you should be incorporating disaster recovery into those deployments. As you add, as things change, disaster recovery shouldn’t be an afterthought. It should be part of everything you do. And writing it down, keeping runbooks, and keeping everything up to date is extremely helpful. Automated clustering can also help because it lets teams test their disaster recovery setups whenever they want and push updates for patching in the background without having to take anything offline. So if you’re thinking of disaster recovery testing as something that you have to do after hours or schedule it in advance, you can build that into kind of daytime hours or production hours in the background. So incorporate it into your operations and not think of it as an afterthought.

Swapnil Bhartiya: Another finding was that 72% of organizations in the survey already use or would consider using HA clustering to streamline patch management. How do you see high availability and cybersecurity coming together?

Natalia Casey: Yeah, I think organizations are recognizing that high availability, security patching, and ransomware recovery, they all serve the same goal, right? Business continuity. So high availability clustering bridges the gap between IT and security, and it can let them roll out critical patches or security updates in the background, again, like, I mentioned before, without having to take systems offline or disrupt their users.

Swapnil Bhartiya: The survey found that only about half of respondents were satisfied with their current HA and DR solutions. What are they running into that a traditional HA or DR approach can’t handle? And if you can talk about how SIOS is taking a different approach for all these, whether it’s traditional workload,

Natalia Casey: I think traditional HA/DR was built for a single data center, and a lot of these things are, you know, we can have one tool for one database and one application. And what SIOS is doing is kind of unifying it. So regardless of where your application resides, which database or which cloud platform, it provides kind of that one view. So you have that one tool to see it, to manage your clusters and to to manage your application availability.

Swapnil Bhartiya: Your report also points to kind of a growing interest in application-aware clustering, especially in mixed Windows and Linux environments. For an IT admin running those systems, what would that actually change in their day-to-day operations and life?

Natalia Casey: Our solution offers application-aware availability, so it takes into account the order of operations and that specific application’s best practices for recovery and replication. So it kind of takes the guesswork out of having to script that all into your deployments. So it’s easy to use and easy to test. So it kind of solves a lot of those issues upfront.

Swapnil Bhartiya: We have been talking about HA, of course we are talking about DR. The whole IT landscape is changing now. AI is also entering the picture. A lot of AI-written code is entering the picture. Agentic AI is making decisions on things. So are tools enough, or do you also feel that within organizations, they also need, as we talk about the cultural change, they also start looking at HA and DR from a different perspective also? Not just that we have the tools in place, but as we talked about, they should be testing it, they should be validating it, they should be, you know, so talk a bit about the importance of processes and cultures within these organizations? Because nothing can ruin your brand when your services are down for an hour.

Natalia Casey: Yeah, I think that HA and DR will solve a different problem with AI. A lot of organizations are looking to AI to solve issues, but I think they’re also not thinking about the issues that AI can create. So I think HA and DR can help with organizations that are now going to leverage agentic AI to sort of build custom deployments to make sure that they are, that they stay online and can run. So AI also makes mistakes. AI is also not foolproof. So you have to, you have to kind of take that into consideration as well.

Swapnil Bhartiya: Natalia, thank you so much for joining me today and walk us through some of the findings of this survey. And also, thank you for sharing your insights on how organizations can improve their HA and DR posture. I really appreciate your time today, and I would love to have you back on the show. Thank you.

Natalia Casey: I appreciate it, Swapnil. Thank you for having me.

Swapnil Bhartiya: And those who are watching, if you want to learn more about SIOS and their high availability and disaster recovery solutions, please visit us.sios.com, and I’ll see you in the next one. Thanks for watching.