Overview
An incident was discovered and some logs were captured. Can you dig through them and uncover details?
We are provided with a packet capture:
dns_tunnel.pcap
The filename itself is an important clue: dns_tunnel.pcap.
The objective is to inspect the network traffic, identify suspicious activity, reconstruct the data that was exfiltrated, and recover the flag.
Prerequisites / Setup
tshark(or Wireshark)xxd/printffor hex decodingdns_tunnel.pcapin the working directory
Recon
Step 1 — Identify the capture file
Goal: Confirm the artifact is a standard PCAP.
file dns_tunnel.pcap
dns_tunnel.pcap: pcap capture file, microsecond ts (little-endian) - version 2.4 (Ethernet, capture length 65535)
Interpretation: The artifact is a standard PCAP packet capture, so Wireshark, tshark, or other packet-analysis tools apply.
Step 2 — Inspect the protocol hierarchy
Goal: Determine what protocols are present.
tshark -r dns_tunnel.pcap -q -z io,phs
Relevant output:
Protocol Hierarchy Statistics
===================================================================
frame frames:3686 bytes:334182
eth frames:3686 bytes:334182
arp frames:240 bytes:10080
ip frames:3446 bytes:324102
udp frames:1366 bytes:126163
dns frames:1306 bytes:120763
ntp frames:60 bytes:5400
tcp frames:2080 bytes:197939
tls frames:511 bytes:111805
data frames:9 bytes:1894
There are:
- 3,686 total Ethernet frames
- 1,306 DNS frames
- 2,080 TCP frames
- 511 TLS frames
- Various ARP and NTP traffic
Interpretation: The large amount of DNS traffic is immediately interesting because the challenge is called dns_tunnel.pcap — DNS may be the exfiltration channel.
Step 3 — Inspect DNS queries
Goal: Extract DNS query names and spot anomalies.
tshark -r dns_tunnel.pcap -Y 'dns.qry.name' \
-T fields -e frame.number -e dns.qry.name
Most of the traffic looks completely normal:
outlook[.]office365[.]com
api[.]slack[.]com
drive[.]google[.]com
cdn[.]jsdelivr[.]net
s3[.]amazonaws[.]com
fonts[.]googleapis[.]com
zoom.us
github.com
graph[.]microsoft[.]com
teams[.]microsoft[.]com
This establishes a background of apparently legitimate DNS traffic.
However, there are three unusual queries:
000[.]723030747b646e735f74756e6e336c73[.]relay[.]blackgate-watch[.]net
001[.]5f63347272795f6d3072335f7468346e[.]relay[.]blackgate-watch[.]net
002[.]5f717533723133737d[.]relay[.]blackgate-watch[.]net
These are clearly different from the normal DNS traffic.
Exploitation
Step 4 — Identify the suspicious domain
Goal: Confirm the domain used for exfiltration.
The suspicious domain is:
relay[.]blackgate-watch[.]net
This is particularly interesting because the challenge itself is named BlackGate Watch.
The three queries have the following structure:
000.<payload>.relay[.]blackgate-watch[.]net
001.<payload>.relay[.]blackgate-watch[.]net
002.<payload>.relay[.]blackgate-watch[.]net
This strongly resembles a DNS tunneling mechanism.
The numerical prefixes:
000
001
002
are likely sequence numbers. The long middle labels are likely encoded data.
Step 5 — Extract the payload
Goal: Confirm the payloads are hex-encoded.
The three suspicious payloads:
723030747b646e735f74756e6e336c73
5f63347272795f6d3072335f7468346e
5f717533723133737d
They contain only hexadecimal characters:
0-9
a-f
This is a strong indication that the data is hexadecimal encoded.
Step 6 — Decode the chunks
Goal: Decode each hex chunk.
First chunk:
printf '%s' '723030747b646e735f74756e6e336c73' | xxd -r -p
r00t{dns_tunn3ls
Second chunk:
printf '%s' '5f63347272795f6d3072335f7468346e' | xxd -r -p
_c4rry_m0r3_th4n
Third chunk:
printf '%s' '5f717533723133737d' | xxd -r -p
_qu3r13s}
Step 7 — Reassemble the chunks
Goal: Order by sequence number and concatenate.
The sequence numbers tell us the correct order:
000 → first chunk
001 → second chunk
002 → third chunk
r00t{dns_tunn3ls
_c4rry_m0r3_th4n
_qu3r13s}
Concatenating the chunks:
r00t{REDACTED}
Flag
r00t{REDACTED}
Recovered by decoding and reassembling the three hex-encoded DNS tunneling payloads.
Lessons / Takeaways
- A DNS hostname can be used as a transport mechanism — inspect unusual subdomains and long middle labels.
- Hex encoding allows arbitrary text to be transported through a DNS label.
- Sequence numbers in query names let you reassemble split payloads in order.
- Compare against a background of legitimate traffic: the three suspicious queries stood out against normal SaaS/CDN DNS queries.
Mitigations
- Monitor DNS query volume and label length/entropy per domain; long hex-like labels on a single dedicated domain are a tunneling indicator.
- Restrict outbound DNS to trusted resolvers and log query names for analysis.
- Use DNS tunneling detection (e.g., unusual qname length, high NXDOMAIN rates, single-client query bursts).
Techniques
References
- libpcap file format: https://wiki.wireshark.org/Development/LibpcapFileFormat

