Detecting DNS Exfiltration with DNS Query Analysis
Overview
DNS exfiltration exploits the Domain Name System as a covert channel to extract data from compromised networks. Attackers encode stolen data into DNS query names (subdomains) or DNS response records (TXT, CNAME, NULL), bypassing traditional security controls that typically allow DNS traffic unrestricted. Tools like iodine, dnscat2, and dns2tcp enable full TCP tunneling over DNS. Detection requires analyzing DNS query patterns for anomalies including excessive query length, high entropy subdomain strings, abnormal query volumes to single domains, and oversized TXT record responses. This skill covers building a comprehensive DNS exfiltration detection capability using passive DNS analysis, statistical methods, and machine learning approaches.
When to Use
- When investigating security incidents that require detecting dns exfiltration with dns query analysis
- When building detection rules or threat hunting queries for this domain
- When SOC analysts need structured procedures for this analysis type
- When validating security monitoring coverage for related attack techniques
Prerequisites
- Access to DNS query logs (passive DNS capture, DNS server logs, or PCAP)
- Zeek, Suricata, or tcpdump for DNS traffic capture
- Python 3.8+ with scipy, numpy, pandas, and scikit-learn
- SIEM platform for alert correlation
- Baseline of normal DNS traffic patterns for the environment
Core Concepts
DNS Tunneling Mechanics
DNS exfiltration encodes data in different parts of DNS messages:
Outbound (Query-based exfiltration):
Encoded data as subdomain labels:
dGhlIHNlY3JldCBkYXRh.exfil.attacker.com
[base64-encoded data].[tunnel domain]
Query types used: A, AAAA, CNAME, MX, TXT, NULL
Inbound (Response-based command channel):
TXT records carry encoded commands/data in responses
CNAME records chain encoded data through multiple labels
NULL records carry arbitrary binary data
Detection Indicators
| Indicator | Normal DNS | DNS Tunneling |
|---|
| Subdomain length | 5-20 chars | 40-253 chars |
| Label count | 2-4 labels | 5-10+ labels |
| Shannon entropy | 2.5-3.5 bits | 4.0-5.5 bits |
| Query volume (per domain) | Variable | 100s-1000s/min |
| TXT response size | < 100 bytes | 200-4000+ bytes |
| Unique subdomains | Low | Very high |
| Query type distribution | Mostly A/AAAA | Heavy TXT, NULL, CNAME |
Common Tunneling Tools
| Tool | Protocol | Encoding | Detection Difficulty |
|---|
| iodine | IP-over-DNS | Base32/Base64/Raw | Medium |
| dnscat2 | TCP-over-DNS | Hex encoding | Medium |
| dns2tcp | TCP-over-DNS | Base64 | Medium |
| DNSExfiltrator | Custom | Base64 | Low |
| Cobalt Strike DNS | C2 over DNS | Custom encoding | High |
Workflow
Step 1: Capture DNS Traffic
Using Zeek:
zeek -i eth0 -C base/protocols/dns
zeek -r traffic.pcap base/protocols/dns
Using tcpdump:
tcpdump -i eth0 -w dns_capture.pcap port 53
tcpdump -i eth0 -w large_dns.pcap 'port 53 and greater 512'
Using Suricata:
outputs:
- eve-log:
types:
- dns:
query: yes
answer: yes
formats: [detailed]
Step 2: Analyze Query Characteristics
Python script for DNS exfiltration detection:
"""DNS Exfiltration Detector - Analyzes DNS logs for tunneling indicators."""
import json
import math
import re
import sys
from collections import defaultdict
from datetime import datetime, timedelta
import pandas as pd
def calculate_entropy(domain: str) -> float:
"""Calculate Shannon entropy of a string."""
if not domain:
return 0.0
freq = defaultdict(int)
for char in domain:
freq[char] += 1
length = len(domain)
entropy = -sum(
(count / length) * math.log2(count / length)
for count in freq.values()
)
return entropy
def extract_subdomain(query: str) -> str:
"""Extract subdomain portion from FQDN."""
parts = query.rstrip('.').split('.')
if len(parts) > 2:
return '.'.join(parts[:-2])
return ''
def get_base_domain() -> :
parts = query.rstrip().split()
(parts) >= :
.join(parts[-:])
query
() -> :
b64_chars = ()
(s) < :
char_ratio = ( c s c b64_chars) / (s)
char_ratio > calculate_entropy(s) >
() -> :
hex_chars = ()
(s) < :
clean = s.replace(, ).replace(, )
(c hex_chars c clean) (clean) % ==
:
():
.domain_stats = defaultdict(: {
: ,
: (),
: ,
: ,
: defaultdict(),
: (),
: ,
: ,
: [],
})
.thresholds = {
: ,
: ,
: ,
: ,
: ,
: ,
: ,
: ,
}
():
base_domain = get_base_domain(query)
subdomain = extract_subdomain(query)
stats = .domain_stats[base_domain]
stats[] +=
stats[].add(subdomain)
stats[] += (subdomain)
stats[] += calculate_entropy(subdomain)
stats[][qtype] +=
stats[].add(src_ip)
stats[] :
stats[] = timestamp
stats[] = timestamp
qtype (, ) response_size > :
stats[].append(response_size)
():
alerts = []
domain, stats .domain_stats.items():
stats[] < .thresholds[]:
unique_count = (stats[])
avg_length = stats[] / stats[]
avg_entropy = stats[] / stats[]
unique_ratio = unique_count / stats[]
txt_queries = stats[].get(, ) + stats[].get(, )
txt_ratio = txt_queries / stats[]
score =
indicators = []
avg_length > .thresholds[]:
score +=
indicators.append()
avg_entropy > .thresholds[]:
score +=
indicators.append()
unique_ratio > .thresholds[]:
score +=
indicators.append()
txt_ratio > .thresholds[]:
score +=
indicators.append()
unique_count > .thresholds[]:
score +=
indicators.append()
encoded_count = (
sd (stats[])[:]
is_base64_like(sd) is_hex_encoded(sd)
)
encoded_count > :
score +=
indicators.append()
score >= :
duration = (stats[] - stats[]).total_seconds() stats[] stats[]
alerts.append({
: domain,
: (score, ),
: stats[],
: unique_count,
: (avg_length, ),
: (avg_entropy, ),
: (unique_ratio, ),
: (txt_ratio, ),
: (stats[]),
: duration,
: indicators,
})
(alerts, key= x: x[], reverse=)
():
(log_path, ) f:
line f:
line.startswith():
fields = line.strip().split()
(fields) < :
:
ts = datetime.fromtimestamp((fields[]))
src_ip = fields[]
query = fields[]
qtype = fields[]
.process_query(ts, src_ip, query, qtype)
(ValueError, IndexError):
():
(log_path, ) f:
line f:
:
event = json.loads(line)
event.get() != :
dns = event.get(, {})
ts = datetime.fromisoformat(event[].replace(, ))
src_ip = event.get(, )
query = dns.get(, )
qtype = dns.get(, )
.process_query(ts, src_ip, query, qtype)
(json.JSONDecodeError, KeyError, ValueError):
():
detector = DNSExfiltrationDetector()
log_file = sys.argv[] (sys.argv) >
log_file.endswith():
detector.process_eve_json(log_file)
:
detector.process_zeek_dns_log(log_file)
alerts = detector.analyze()
alerts:
()
()
()
alert alerts:
severity = alert[] >= alert[] >=
()
()
()
()
()
()
()
:
()
__name__ == :
main()
Step 3: Deploy Suricata Rules for DNS Exfiltration
# Detect long DNS queries (potential tunneling)
alert dns $HOME_NET any -> any 53 (msg:"DNS Exfiltration - Excessive query length"; dns.query; content:"."; pcre:"/^.{60,}/"; threshold:type both,track by_src,count 20,seconds 60; classtype:bad-unknown; sid:3000001; rev:1;)
# Detect high-entropy DNS subdomain
alert dns $HOME_NET any -> any 53 (msg:"DNS Exfiltration - High entropy subdomain"; dns.query; pcre:"/^[a-zA-Z0-9+\/=]{30,}\./"; threshold:type both,track by_src,count 10,seconds 60; classtype:bad-unknown; sid:3000002; rev:1;)
# Detect large TXT record responses
alert dns any 53 -> $HOME_NET any (msg:"DNS Exfiltration - Large TXT response"; content:"|00 10|"; byte_test:2,>,400,0,relative; classtype:bad-unknown; sid:3000003; rev:1;)
# Detect NULL record queries (used by iodine)
alert dns $HOME_NET any -> any 53 (msg:"DNS Exfiltration - NULL record query (iodine indicator)"; content:"|00 0a|"; classtype:bad-unknown; sid:3000004; rev:1;)
# Detect dnscat2 traffic pattern
alert dns $HOME_NET any -> any 53 (msg:"DNS Exfiltration - dnscat2 indicator"; dns.query; content:"dnscat"; nocase; classtype:trojan-activity; sid:3000005; rev:1;)
Step 4: SIEM Detection Rules
Splunk SPL query for DNS exfiltration:
index=dns sourcetype=zeek:dns
| eval subdomain=mvindex(split(query,"."),0)
| eval subdomain_len=len(subdomain)
| eval label_count=mvcount(split(query,"."))
| stats count as query_count,
dc(subdomain) as unique_subdomains,
avg(subdomain_len) as avg_sub_len,
values(src_ip) as source_ips
by query_domain
| where query_count > 100 AND avg_sub_len > 30 AND unique_subdomains > 50
| eval risk_score = case(
avg_sub_len > 50 AND unique_subdomains > 200, "Critical",
avg_sub_len > 40 AND unique_subdomains > 100, "High",
avg_sub_len > 30 AND unique_subdomains > 50, "Medium",
true(), "Low")
| sort -query_count
| table query_domain risk_score query_count unique_subdomains avg_sub_len source_ips
Response Actions
- Block the tunnel domain at DNS resolver and firewall level
- Isolate the source host from the network for forensic investigation
- Capture full PCAP of the DNS traffic for evidence preservation
- Identify exfiltrated data by decoding captured DNS queries
- Check for persistence mechanisms on the compromised host
- Update blocklists with identified C2 domains and infrastructure
Best Practices
- DNS Logging - Enable full DNS query and response logging at resolvers and network level
- Internal DNS Only - Force all DNS through internal resolvers; block direct external DNS (port 53)
- Response Policy Zones - Deploy RPZ feeds to block known tunneling domains
- Baseline First - Establish normal DNS query patterns before setting detection thresholds
- TXT Record Monitoring - Pay special attention to TXT and NULL record queries
- Encrypted DNS Awareness - Monitor for DoH/DoT usage that may bypass DNS inspection
References
