This article is more than 1 year old
IDS users swamped with false alerts
Four in five are duds
The number of redundant alarms and false positives generated by Intrusion Detection Systems (IDS) has come under fire from users attending an event designed to raise awareness about the technology.
Users attending an 'ABCs of IDS' event at London's City University yesterday said more the 80 per cent of the alerts they received were false, with one citing 60 alerts he had received about non-existent problems that morning at 0300.
IDS systems, which act as a kind of burglar alarm to for hacker attacks, provide "defence in depth" from hack attacks by providing alerts about suspicious activity via sensors on a network or host machine. The technology is reactive in nature, meaning human intervention is needed, because if an IDS system shut off traffic or stopped a service automatically this could cause more problems than it solves.
Representation from Cisco Systems, Intrusion.com, NFR Security, Top Layer and managed services firm proseq at the event conceded false alerts and redundant alarms (such as Apache-targeted hacks on firm with no such servers) were a serious problem. However there was a marked absence of ideas of how about the technology could be fine-tuned to minimise the problem, with vendors emphasising the fact that systems need to be part of an overall security policy, which receives adequate resources.
Part of the problem seems to be that business managers buy IDS systems (often on the advice of auditors or consultants) without committing to the people and resources needed to make the technology work, or having a managed services firm maintain an installation.
The concern is that adopters of the technology will fail to maintain it or simply leave it to gather dust as overworked admins get bombarded with false alarms.
In recent weeks we've spoken to two service providers, COLT Telecom and Data Return, who both told us customers request the installation of IDS systems in their datacentre but then subsequently fail to monitor the alerts generated.
Speaking at a recent Black Hat conference, Nicolas Fischbach, senior IP and security engineer at COLT Telecom, said IDS systems if hosting environments generate "thousands of alerts but no one watches them."
This can generate a false sense of security, he warned.
Users at last night's conference echoed these concerns and one summed up the feelings of the meeting by saying "installation of IDS systems is only 10 per cent of the solution".
Brian Milnes, general manager in Northern Europe of security tools firm Intrusion.com, said buying a IDS systems was like "buying a Christmas puppy" because both needed attention. ®
Stealth encoding bypasses IDS protection
The Cisco Intrusion Undetection System
Hacking activity at all-time high
Denial of service warning for network security tool
Different approach to intrusion detection touted