What the online world can learn from recent on prem Exchange Server challenges

It has been a pretty challenging few days for those that still manage and maintain on premises Exchange servers thanks to:

HAFNIUM targeting Exchange Servers with 0-day exploits

Throughout which I’ve seen a lot of smug cloud administrators wondering why people still bother with on premises. I think a better use of their energies would be to look at the current situation and learn from it rather than allocating it to self righteousness.

The cloud is a shared responsibility model. This means that both Microsoft and end user now responsible for the security of cloud infrastructure. Luckily, these recent Exchange issues have largely fallen to Microsoft when it comes to the cloud. Where there is room to learn for the rest of us, is in the response to the situation from those battling to contain it.

From everything I have seen online in regards to the HAFNIUM issue, what I find most interesting is the lack of a response plan. Technically, administrators can follow directions, run scripts, patch systems pretty well. However, most seem totally unprepared for this kind of situation, especially at scale. That’s what worries me the most. Why? Because challenges in the cloud can easily be of the same scale and impact.

There have been plenty of examples when services like Azure AD or Exchange Online have been unavailable, but when they have, I’ve seen the same level of, dare I say, panic. Because systems work 99.99% or more of the time ‘on average’, a large amount of complacency begins to creep into the system, especially those charged with maintaining these systems. Thoughts of disaster recovery and outage impact get put on the back burner and never really addressed because there are always ‘higher’ priorities.

What worries me is the dependency we have built into our modern lives, business and economy, to the point where most cannot function if their phones run out of charge. What worries me when I look at the response I see to broad security challenges in IT is simply the lack of a credible contingency plan. A check list of what to do, if you like. Of course, you can’t have a plan for every contingency but some semblance of a plan is better than no plan at all surely?

In the end it comes down to risk analysis. When the sun is shining, risk analysis is the furthest thing from people minds. This however, is exactly the time that it should be a priority because developing a strategy in midst of a crisis does not generally lead to the best outcome. You want to have a checklist of what to do, well in advance of whenever you may need it.

Even though the systems I work with are cloud based are immune from the HAFNIUM (it appears at least), that doesn’t stop me learning from how the unfortunate are dealing with it. I’m watching, learning and preparing, because as the saying goes, “When did Noah build the Ark?”

Before it rained.

Before it rained.