Abstract

Deception techniques - if deployed well - can be very effective for organizations to improve network defense and can be a useful arsenal for blue teams to detect attacks at very early stage of cyber kill chain. But the challenge we have seen is deploying, managing and administering decoys across large networks is still not easy and becomes complex for defenders to manage this over time. Although there are a lot of commercial tools in this space, we haven’t come across open source tools which can achieve this.

With this in mind, we have developed DejaVu which is an open source deception framework which can be used to deploy across the infrastructure. This could be used by the defender to deploy multiple interactive decoys (HTTP Servers, SQL, SMB, FTP, SSH, client side – NBNS) strategically across their network on different VLAN’s. To ease the management of decoys, we have built a web-based platform which can be used to deploy, administer and configure all the decoys effectively from a centralized console. Logging and alerting dashboard displays detailed information about the alerts generated and can be further configured on how these alerts should be handled. If certain IP’s like in-house vulnerability scanner, SCCM etc. needs to be whitelisted, this can be configured which effectively would mean very few false positives.

Alerts only occur when an adversary is engaged with the decoy, so now when the attacker touches the decoy during reconnaissance or performs authentication attempts this raises a high accuracy alert which should be investigated by the defense. Decoys can also be placed on the client VLAN’s to detect client side attacks such as responder/LLMNR attacks using client side decoys. Additionally, common attacks which the adversary uses to compromise such as abusing Tomcat/SQL server for initial foothold can be deployed as decoys, luring the attacker and enabling detection.