dcsimg
Linux Today: Linux News On Internet Time.




More on LinuxToday


Linux Journal: PortSentry

Jul 14, 2001, 14:00 (5 Talkback[s])
(Other stories by Anthony Cinelli)

A handy piece of software for detecting and dealing with port scans is PortSentry. This article looks into installation and basic configuration of this package:

"As any administrator knows, a successful network rollout begins and ends with security. No matter how much money is spent on a system with the latest and greatest hardware and software, the system can be rendered worthless if its security is compromised. Unfortunately, keeping up with system security can be tedious. Administrators must stay aware of updates to software as well as the latest system compromise techniques. Due to this difficult task, system security is often not maintained and is lacking in many areas. This is illustrated by the increased number of reports that entail system compromise. This dilemma changed for me when I discovered the freeware tools offered by Psionic Software, Inc. called PortSentry and Logcheck. Within minutes, these tools can be installed and configured to improve system security dramatically.

Once a host is targeted by an attacker, a port scan is almost always performed. The port scan is done to expose all services available on the target host and to provide a starting point for break-in attempts. PortSentry detects such scans by monitoring the unused ports on the host. Upon a connection attempt to one of the unused ports, PortSentry is alerted and has the ability to issue a number of commands in response to the scan. The commands issued are configured by the administrator within a configuration file. Although any command may be used, the most helpful is one in which the IP address of the attacker's host is essentially "black holed" by issuing a routing command that denies all traffic from that address. The violation and corresponding action taken by PortSentry are recorded in the system log. Using another Psionic utility, Logcheck, these security alerts are e-mailed to an administrator at designated intervals. Thus, the host is now capable not only of retaliating against a potential break-in attempt automatically, but also of notifying the administrator of the occurrence."

Complete Story

Related Stories: