Recognize attacks on network infrastructure
- On-path interception
- An attacker positioned between parties can observe traffic and may alter, inject, or block messages. An attacker who obtains a reusable session identifier may replay it to impersonate the user. Man-in-the-browser malware can intercept or alter transactions inside a compromised browser, even when traffic between the browser and server uses TLS.
- DNS poisoning and hijacking
- Cache poisoning introduces false records into a DNS cache. DNS hijacking changes how names are resolved, for example by modifying a device’s resolver settings or taking control of authoritative DNS records. Either can steer a user to the wrong IP address.
- Distributed denial of service (DDoS)
- Traffic from many sources can consume network, system, or application capacity. A botnet can send requests directly. In a reflection attack, services send replies to a spoofed source address belonging to the victim. Amplification occurs when those replies are larger than the requests.
- Layer 2 attacks
- ARP poisoning makes a local IPv4 host associate an IP address with the wrong MAC address. MAC cloning impersonates another device’s MAC address and can cause a switch to relearn it on the wrong port. MAC flooding pressures the switch’s address table. Depending on the switch and its configuration, some unknown unicast traffic may then be flooded within the VLAN.
- Credential relay
- An attacker forwards a live authentication exchange to another service so it accepts the victim’s authentication. Relay can work without learning the password. Protocol protections such as signing or channel binding can help where supported.
Legitimate automation tools can also execute malicious instructions. PowerShell runs administration tasks and scripts. VBA automates tasks with Office macros. Python runs on multiple platforms, and Bash interprets shell commands on Unix-like systems. Review the script’s source, parent process, account, and follow-on activity. Use signed macros or execution restrictions where appropriate and keep scripting telemetry available for investigation.
CISA · Mitigate DNS Infrastructure Tampering (PDF) · CISA and partners · Understanding and Responding to DDoS Attacks (PDF)
Use complementary monitoring methods
- Signature-based
- Matches an observed event to a known pattern, such as traffic to a documented malicious destination.
- Anomaly-based
- Compares current activity with a baseline, such as normal traffic volume for that service and time of day.
- Behavior-based
- Looks for actions or sequences linked to an attack, such as an Office process launching a script interpreter that then opens an outbound connection.
- Heuristic
- Uses rules of thumb or scoring to identify suspicious combinations, such as a new sender domain combined with an unusual attachment type.
Use scans to identify exposed services and weaknesses. Generate alerts for events that warrant action and reports for trends or recurring issues. Triage alerts, quarantine messages or isolate systems where appropriate, and tune detections using confirmed incidents and legitimate activity. Archive relevant records with synchronized timestamps, access controls, and a retention schedule that supports investigations.
Pick telemetry that answers the question
- Packet analysis. Examines captured packets for protocol details, DNS queries, unusual endpoints, or recurring connections that may indicate command and control. Encryption and capture location affect what is visible.
- Flow analysis. Summarizes conversations using fields such as endpoints, ports, protocol, timing, and byte and packet counts. Cisco NetFlow exports flow records from observed IP traffic. Sampling is optional in some deployments, and collection does not depend on ICMP Echo requests.
- DLP. Data loss prevention tools identify sensitive content through labels, patterns, or other signals and can alert, quarantine, or block transfers that violate policy. Coverage depends on which channels and systems the tool monitors.
- SNMP traps and log aggregation. A managed device can send an unsolicited SNMP trap to a manager when an event occurs. Log aggregation brings records from different sources together for search and correlation.
- SCAP. The Security Content Automation Protocol standardizes how tools describe software flaws and configuration checks. Compatible scanners can use machine-readable benchmarks to assess systems against policy.
- SIEM and SOAR. Security information and event management (SIEM) collects and correlates activity for detection and reporting. Security orchestration, automation, and response (SOAR) coordinates response workflows, such as enriching an alert or requesting approval to isolate an endpoint.
Cisco · NetFlow Configuration Guide (PDF) · NIST · SCAP Version 1.4
Monitor email routing and authentication
A mail user agent (MUA) is the app or webmail interface people use to read and send mail. A mail transfer agent (MTA) accepts and routes messages. MTAs add trace information such as Received fields. Inspect the headers added by systems you trust, including their authentication results. An untrusted sender can insert forged header fields.
- SPF
- Checks whether the sending host is authorized for the domain used in the SMTP envelope. That domain can differ from the visible From address.
- DKIM
- Verifies a domain’s signature over selected message headers and the signed body content.
- DMARC
- Checks whether at least one passing SPF or DKIM result aligns with the visible From domain. It also supports domain policy and reporting. A look-alike domain can still pass authentication for its own domain.
A secure email gateway (SEG) can sit in the mail path and inspect messages for phishing, malicious links and attachments, and policy violations. It can quarantine suspicious messages and provide evidence for investigation. Authentication results and gateway verdicts each answer specific questions about a message.
Practice case
An invoice email and an unusual connection pattern. A finance employee reports an unexpected invoice from billing@acme-payrnents.example. The supplier’s domain on file is acme-payments.example. The mail gateway records a DMARC pass. After the message arrives, DNS logs record the employee’s workstation looking up the look-alike domain. Flow records then show brief connections to the returned IP address roughly once a minute.
Why can the message pass DMARC? What do DNS and flow records establish? Which evidence would you check to determine what launched the connections and whether data left the workstation?
Review a possible answer
The look-alike domain can publish its own authentication records and pass DMARC for its own From domain. Compare the address with the supplier record and examine the gateway’s trusted authentication results. A DNS lookup shows a resolution attempt. Flow records can show connection timing and volume. A periodic pattern is consistent with automated activity and deserves investigation.
Correlate timestamps with browser records and endpoint process telemetry to identify what launched the connections. Review proxy, DLP, or authorized packet records for evidence of transferred content. Preserve the message and its headers, and follow the response plan if the combined evidence supports quarantining the message or isolating the host.
Read and watch
- Compare DDoS mechanisms and response options. CISA and partners · Understanding and Responding to DDoS Attacks (PDF)
- Review sender alignment, policy, and reporting. RFC 9989 · DMARC
- Watch how a packet capture is collected and filtered. Video · Wireshark Tutorial for Beginners