Computer security using virus probing

DETAILED DESCRIPTION The present invention provides a technique for determining whether particular clients within a computer network are universally configured in accordance with the desired security features of the computer network.
In accordance with the invention, a probe is randomly inserted within incoming files in the computer network.
Illustratively, the insertion of probes occurs in a firewall which separates the computer network from other networks.
The probe, in accordance with an embodiment of the invention, is configured as a function of a particular execution task, e.
g.
a known virus, such that in a properly configured client the probe will not execute and the firewall does not detect a security breach.
However, if the client is misconfigured, i.
e.
, is not in compliance with the standard network security measures, the probe will execute and trigger a security alert in the firewall indicating that the client is vulnerable to a security breach.
Advantageously, a network security administrator can take appropriate action to correct those clients which are misconfigured.
In preferred embodiments of the invention, the probe is configured as a virus probe in the form of a trojan horse which, if executed, on a client will launch a signal back to the firewall indicating that the client is misconfigured.
In further embodiments of the invention, the signal back to the firewall is a User Datagram Protocol ("UDP") packet.
In accordance, with a further embodiment of the invention, the virus probe is inserted upon a first Internet access from a particular IP address or browser type, and thereafter virus probes are inserted at random intervals.