It is a pcap capture with all the traffic (background, normal and botnet)
This pcap file was not made public because it contains too much private information about the users of the network.
This file was captures on the main router of the University network.
Capture with only the botnet traffic. It is made public.
This file was captured on the interface of the virtual machine being infected.
This file has the netflows generated by a unidirectional argus. The labels were assigned as this:
- First put Background to all the flows. - Put LEGITIMATE to the flows that match some filters. - Put Botnet to the flows that come to or from the infected IP addresses
- Infected hosts - 220.127.116.11: Windows XP (English version) Name: SARUMAN (Label: Botnet) (amount of bidirectional flows: 20941) - Normal hosts: - 18.104.22.168 (amount of bidirectional flows: 8960, Label: Normal-V42-Stribrek) - 22.214.171.124 (amount of bidirectional flows: 25, Label: Normal-V42-Grill) - 126.96.36.199 (amount of bidirectional flows: 71, Label: CVUT-WebServer. This normal host is not so reliable since is a webserver) - 188.8.131.52 (amount of bidirectional flows: 23, Label: CVUT-DNS-Server. This normal host is not so reliable since is a dns server) - 184.108.40.206 (amount of bidirectional flows: 3, Label: MatLab-Server. This normal host is not so reliable since is a matlab server)
Please note that the labels of the flows generated by the malware start with "From-Botnet". The labels "To-Botnet" are flows sent to the botnet by unknown computers, so they should not be considered malicious perse. Also for the normal computers, the counts are for the labels "From-Normal". The labels "To-Normal" are flows sent to the botnet by unknown computers, so they should not be considered malicious perse.
We started the capture of the traffic of the whole university. Whitout the botnet.
We limited the bandwith of the experiment to 20kbps in the output of the bot.
We started the neris malware.
We stopped the botnet.
We stopped both captures.
These files were generated in the Stratosphere Lab as part of the Malware Capture Facility Project in the CVUT University, Prague, Czech Republic. The goal is to store long-lived real botnet traffic and to generate labeled netflows files. Any question feel free to contact us: Sebastian Garcia: email@example.com You are free to use these files as long as you reference this project and the authors as follows: Garcia, Sebastian. Malware Capture Facility Project. Retrieved from https://stratosphereips.org