Abhinav Saxena wrote this post as a project summary of his GSoC2018 experience.

What did we achieve?

The following features and changes were implemented:

• Migration of the codebase from Python 2.7 to Python 3.5 (issue #358, code: #374)
• Implementation of FTP (RFC 959) and TFTP (RFC 1350) protocol stacks based on gevent (issue #352, code: ftp and tftp)
• Implementation of an abstract filesystem that proxies and wraps an actual file system by providing os.* wrappers (code: #375 and #382)
• Wrote 123 unit tests and refactored all existing 44 unit tests, increasing coverage from 44% to 72% at the time of this writing  (code: #374, #375 and #382)
• Bug fixes and refactoring of the existing BACnet and IPMI protocol stacks (issue #341, code #382)
• Bug fixes in auxiliary Docker files (issue: #378, code: #380 and  #392)
• Refactoring of an existing telnet library to be compatible to the Conpot codebase (issue #285, code: mushorg/telnetsrvlib)
• Wrote an internal interface implementation that introduces a decorator, allowing protocol servers to interact more deeply with each other.  (issue #259, code #375)
• Helping users with issues and pull request reviews: link

All commits can be seen here and here.

The Conpot development team is proud to announce the 0.5.0 release. Highlights of this release are the support for two new protocols and one additional device. Peter Soóky did a major contribution with support for the BACnet protocol, which is used for building automation and control networks, and support for IPMI, which is used an interface to a computer subsystem that provides management and monitoring capabilities independently of the host system’s CPU, firmware and operating system (consider the insights you can get from someone exploiting this). As mentioned in an earlier blog post, we also added support to emulate a Guardian AST device. This is based on the research from Kyle Wilhoit and Stephen Hilt.
Another goal of this release was to improve the ease of deployment. Therefore we added a Docker container template. Thanks to our contributors, we also have documentation on how to run Conpot on CentOS.
To avoid some easy fingerprinting, we added the feature to modify the MAC address of the interface Conpot is listening on. So now your hardware address can match the device manufacturer you are intending to emulate.
As with every other release, we tried to improve our test coverage and code quality in order to increase the honeypots stability.

The Conpot team is following closely the latest developments in Honeypot research and the methods and technologies used. If you look at the topics presented on security conferences, you might have also noticed an increased interest in ICS security and honeypot technologies in the last two years. One presentation from this years Blackhat’15 conference caught my attention also knowing previous research done by Kyle and Stephen: “The little pump gauge that could: Attacks against gas pump monitoring systems” [link] If you are interested in their findings, I recommend their white paper: “The GasPot Experiment: Unexamined Perils in Using Gas-Tank-Monitoring Systems“ [link, pdf] by Kyle Wilhoit and Stephen Hilt from Trend Micro’s Forward-Looking Threat Research team.

Lukas Rist is a software engineer with Blue Coat Norway where he develops behavioral malware analysis systems. In his spare time, he works on web application and ICS/SCADA honeypots and botnet monitoring tools under the umbrella of the Honeynet Project where he is also a Director. He recently developed an interest in deployment automation, ephemeral file systems and exotic industrial communication protocols.

1) What was your motivation to enter Information Security field, and who inspired and helped you along the way?

The team working on the ICS/SCADA honeypot Conpot, just merged in a more mature support for STIX (Structured Threat Information eXpression) formatted reporting via TAXII (Trusted Automated eXchange of Indicator Information) into the master branch on Github.

STIX allows us to represent event sessions captured by the honeypot in a structured format, which eases the integration of Conpot into existing consumer (e.g. SIEM) infrastructures.

By transforming an arbitrary honeypot event into a schema defined format, we are able to communicate an incident in a language, which is also understandable by someone not trained in interpreting industrial protocol messages.