Have recent vulnerabilities changed how you approach infrastructure security?

November 17, 2026 5 answers
Asked by _w52~ferrulecanoe.6

Are the recent high-profile vulnerabilities pushing teams toward more preventative infrastructure security, or is the approach still mostly patching and applying mitigations?

I'm curious what people would prioritize, from isolation and least privilege to runtime monitoring and better visibility.

Answers

Mirevalo-2s

I'd put more weight on least privilege and workload isolation. Patching still matters, but there may not always be time to deploy a fix before exploitation starts. Limiting what a compromised service can access gives you another layer of protection.

Runtime monitoring can help too, especially when it looks for unexpected process behavior rather than relying only on known signatures. The tricky part is keeping the alerts useful enough that people don't tune them out.

_w52~ferrulecanoe.6:

What sort of behavior would you start monitoring?

Reply to this answer
7t~norlivo_536

Inventory and ownership come first for me. If you can't quickly find where affected software is running, identify who owns it, and work out how to roll out a fix safely, the rest of the response gets messy fast.

That doesn't mean ignoring preventative controls. It means making sure you can actually find and update the things those controls are supposed to protect.

Reply to this answer
_u72!ferrulegizmo.5

For virtual machines, look at what shares a virtualization cluster, not just what shares a network. Moving public-facing workloads onto a separate perimeter cluster could limit the impact of a guest-to-host escape on internal workloads.

It's a change worth considering during a hardware refresh, when the infrastructure is already being planned.

Reply to this answer
_m29!bravexa.4

There are some fairly practical hardening steps before adding a noisy monitoring system. Use minimal images, make the root filesystem read-only where the workload supports it, and consider mounting temporary directories with noexec. Those controls have limits, but they can reduce opportunities for unwanted execution.

Test the application and maintenance jobs against those restrictions first. A control that breaks normal operations won't stay enabled for long.

Reply to this answer
_k73~ferrinook.8

I'm not convinced the underlying strategy has necessarily changed. Some vulnerabilities get much more attention than others, and a rising count of published CVEs alone doesn't tell you how much the risk has changed. More organizations are now authorized to assign CVEs, which also affects the count.

Reply to this answer

Add your answer

Answers are reviewed before publication.

Related Discussions