Operating System Security
Processes, users, file permissions, privileges, patching and resource limits decide what a compromised application can do next.
Frame the problem
Security starts with a concrete asset, attacker capability and trust crossing.
Why the system fails
The service runs with broad user, filesystem, capability or resource privileges and relies on application correctness as its only boundary.
The important question is not “what is Operating System Security?” but “which assumption let untrusted data or an over-scoped identity cross process identity → kernel-mediated resources?” Trace the decision at the boundary, then constrain what can happen after the first control fails.
Design the control in layers
Start with the control closest to the interpretation or privilege boundary: Run as a dedicated restricted identity Then add a control that reduces blast radius and telemetry that proves the decision was enforced.
The resulting design is not labelled secure. Record the identified controls, the known failure paths, the remaining exposure, and the evidence you would need during an incident.
| Prevent | Detect | Recover |
|---|---|---|
| Run as a dedicated restricted identity · Remove capabilities and writable paths · Patch and apply CPU, memory and process limits | Unexpected syscalls, children, file access and privilege changes | Contain the affected identity or component, scope impact from audit evidence, and preserve a regression test. |
Key points
- Asset: Host kernel, files, credentials and neighboring workloads.
- Boundary: Process identity → kernel-mediated resources
- Primary control: Run as a dedicated restricted identity
- Detection signal: Unexpected syscalls, children, file access and privilege changes
- Always ask what limits damage when the primary control fails.
Follow the attack
Safe conceptual simulation: capability → missing control → crossed boundary → asset impact.
- 1Attacker starts with: A compromised process or malicious local user.
- 2The service runs with broad user, filesystem, capability or resource privileges and relies on application correctness as its only boundary.
- 3The weak or missing boundary control is crossed: Process identity → kernel-mediated resources
- 4Impact: Host persistence, secret access or denial of neighboring workloads.
- Host persistence, secret access or denial of neighboring workloads.
Defend, detect, recover
One prevention is a single point of security failure. Layer it and make failure observable.
- • Run as a dedicated restricted identity
- • Remove capabilities and writable paths
- • Patch and apply CPU, memory and process limits
- • Unexpected syscalls, children, file access and privilege changes
- • Contain the affected identity or component.
- • Scope access from audit evidence.
- • Fix the boundary and add a regression test.
- • Misconfiguration and new access paths can bypass the intended control.
- • A privileged insider or compromised control plane may still reach the asset.