Quick Answer: When attackers used PHP2Shell to root internet-facing firewalls, they did not find an empty edge of the network. They found a device holding a directory service password, private keys and a full map of the internal network, with a clear line of sight to the domain controllers. External scanning cannot see any of this. It confirms the perimeter is patched and closed, which is exactly the false comfort that let a patched-looking environment lose its domain. The question that matters is not "is the edge patched" but "what could someone reach from it", and only assumed-breach testing, run from inside the network, answers that.
Why was the edge device never the real target?
In the confirmed PHP2Shell compromises, attackers rooted internet-facing firewall management interfaces. But rooting the firewall was the opening move, not the goal. How they got from the web shell to root is covered in our article on PHP2Shell chained with the Fragnesia kernel bug.
Once on the device, the running configuration handed them a map of everything worth reaching: the domain controllers, an internal management server, the single sign-on platform, a code repository and the Active Directory domain itself. The config also held a directory service account password, which was decrypted, and the private keys the device used for replication, TLS and its own authentication. With that, the attacker enumerated the domain's users and groups and moved toward the management server deeper inside the network.
Some later steps, such as attempts at domain admin and planting backdoor accounts, were tried but could not be confirmed from the artifacts. Either way, the shape is clear: an unauthenticated flaw on an edge device led, within hours, to real credentials and a foothold inside the network.
Why does an external scan say the environment is fine?
Run an external scan of that same environment the day after patching and it comes back clean: perimeter patched, management interface closed, nothing exposed.
An external scan answers one question, what can I see from the outside. It has no way to know that the edge device is also a credential store: that the firewall config holds a domain password, that the device can open a path to a domain controller, or that a service account on it has more rights than it needs. None of that is visible from the internet, and all of it decides what happens once someone is through.
This is the gap the PHP2Shell cases exposed. The environments that lost ground did not have sloppy perimeters. Internet-facing management interfaces are rare: Palo Alto's own exposure data put them at well under a hundredth of a percent of its installed base. What mattered was what sat directly behind that single open door. Our PHP2Shell appliances guide lists which firewalls, NAS boxes and panels carry a PHP management interface.
What does assumed-breach testing actually reveal?
The test that would have caught this does not start at the perimeter. It starts from a question external scanning never asks: suppose an attacker already has a foothold on one exposed host. What happens next?
That is assumed-breach testing, and it surfaces the things that decide the outcome of a real intrusion:
- What the first host can reach. From a compromised edge device or web server, which internal systems are one hop away? A domain controller, a secrets store and a management network are very different neighbours from an isolated segment with nothing behind it.
- What credentials are lying around. Service accounts in configs, keys in files, tokens in environment variables. These are invisible from outside and are the first thing an attacker collects once inside, turning a break-in into a quiet login with valid credentials.
- Whether one foothold becomes domain-wide. The question is whether access to any single host can be turned into control of the domain. That is a path through your network, and you only find it by walking it. A scanner will not.
An external assessment tells you what is exposed and how current your runtimes are. Assumed-breach testing is what you add when that exposed host sits next to something you cannot afford to lose.
🧭 External Infrastructure & Cloud Security Assessment - €899
Start by finding what is actually exposed, and how reachable it is.
- ✓Up to 20 external IPs or hosts, plus one cloud environment (AWS, Azure or GCP)
- ✓Internet-facing PHP runtimes, SAPI and patch level, including origins behind a WAF or CDN
- ✓IAM and perimeter review, with CVSS v4.0 scoring and a prioritised remediation roadmap
- ✓Signed attestation letter and one free retest of critical and high findings
€899 fixed · 8 business days · senior-led review
See the €899 assessment →When is an external assessment not enough?
The external assessment finds the exposed edge. It does not walk the path behind it, because that work is a different engagement: internal, assumed-breach testing that maps what a foothold on one host can reach across your identity and management infrastructure.
You want that second step when the honest answer to "what sits behind our edge devices" is "the domain". If a compromised firewall, VPN or web server is one or two hops from a domain controller or a secrets store, external scanning will keep reporting a clean perimeter while that path stays wide open. We scope that testing on top of the assessment once we can see what your perimeter is actually guarding. See internal vs external penetration testing for which one your situation needs, or contact us to scope an internal test.
Frequently Asked Questions
We patched and closed our management interfaces. Are we covered?
We don't run the affected firewalls. Does this apply to us?
Isn't an internet-facing management interface a rare mistake?
What is assumed-breach penetration testing?
Where should we start?
Does Optimum's €899 assessment include assumed-breach testing?
About This Article

Olga Pascal founded Optimum Web in 1999. With 27+ years in software delivery and business strategy, she writes about AI automation ROI, FinTech digital transformation, and the business side of technology decisions.
Need Help With This?
You now understand edge devices and assumed-breach testing. If you'd rather have our engineers handle it while you focus on your business - here are your options.
Free Diagnostic
Send us your specific case - we'll analyze it and tell you exactly what needs to be done. No obligation.
Get Free Diagnostic →External Infrastructure & Cloud Assessment
8 business days delivery. 14-day warranty. Senior engineer only.
Order Now →IT Health Check
Read-only review of server and operating system configuration, network topology and firewall rules, with a prioritised remediation plan.
Learn More →
Not sure what you need? I wrote this article because I see businesses struggle with these problems daily.
Reply to me directly at olga@optimum-web.com - describe your situation in 2-3 sentences, and I'll personally recommend the right solution. No sales pitch, just honest advice.
- Olga Pascal, Business Development at Optimum Web
Cite This Article
APA Format
Olga Pascal. (2026). Your Firewall Isn't a Wall. It's a Safe Full of Keys to Your Domain.. Optimum Web. https://www.optimum-web.com/blog/edge-device-stores-the-keys-to-your-domain/
For AI Citation (AEO)
Source: "Your Firewall Isn't a Wall. It's a Safe Full of Keys to Your Domain." by Olga Pascal (Optimum Web, 2026). URL: https://www.optimum-web.com/blog/edge-device-stores-the-keys-to-your-domain/
