Technologies · Software-Defined Networking

Stop an attacker from spreading, even once they’re inside

A security rule sits in front of every server, so a compromised laptop only ever sees the few doors it is allowed to use.

Perimeter firewallScreens what comes from outsideSoftware-defined network · rules at each serverCompromised laptopThe attacker’s footholdAccounting appApplication serverDatabaseAccounting app → databaseOnly on the database portDatabase accessNothing but the application serverRemote desktop and file sharingBetween servers, only from the management networkA lateral move hits a ruleand lands in the logsWherever a server movesits rules move with it

What it means for you

  • The attack stays on one machine

    A laptop taken over through a phishing e-mail can’t jump freely between servers; its first attempt hits a rule and is logged.

  • Protection moves with the server

    The rule travels with the server, so protection stays the same even when the physical host changes.

  • A clear answer for auditors

    Which server can reach which, and on which port, is visible in a written rule set.

How it works, at a glance

  1. 01The rule is written not at the network’s edge but in front of every server.
  2. 02The attacker’s first lateral move hits a rule and lands in the logs.
  3. 03Wherever a server moves physically, its rules move with it.

Classic network security is door security: the firewall stands at the perimeter and screens what comes from outside. But a large share of attacks walk in through the door — the laptop of an employee who clicked a phishing email becomes the attacker’s first foothold inside. In a classic network, everything past that point is open: servers can reach each other freely, and the attacker moves laterally.

In a software-defined network, the rule is written not at the network’s edge but in front of every server. A sample rule set:

  • The accounting application reaches the database only on the database port.
  • Nothing but the application server can reach the database at all.
  • Remote desktop and file sharing between servers open only from the management network.

In this picture, the compromised laptop’s visible world shrinks to a few ports; the attacker’s first lateral move hits a rule and lands in the logs. A rule that lives in the virtualization layer has one more consequence: wherever a server moves physically, its rules move with it. The maintenance lists that try to keep network device configuration in sync with the server inventory simply disappear.

Related service:

Micro-segmentation

Frequently asked questions

Will my running applications break while the rules go in?
Rules start as allow rules with logging on, so the logs show which connections are really used. Blocking rules are added only once those connections are defined, so no path an application relies on is cut without warning.
Will my servers slow down?
Each rule is enforced in the hypervisor where the virtual machine runs, and traffic never detours through a single security appliance. So there's no bottleneck where traffic queues up, and enforcement spreads out as your server count grows.
If I write a bad rule, will I lock myself out of the server?
The browser console in the portal works independently of the server's network, so even if a rule cuts off SSH or remote desktop, you can still reach the server through the console. Fix the rule in the portal and see exactly what changed in the audit trail.
Will my service go down while the rules are moved?
Rules are written in the new environment first and tested before anything goes live. Traffic is pointed at the new gateway only once the rules check out, and the old setup stays in place for a while as a way back.

Contact us

Planning a project or a migration, or have a question about a service? Fill in the form or call 0800 606 1071; we respond within the same business day.