Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Network forensics reconstructs an intrusion from traffic records rather than from the endpoints, using packet captures, flow data, and protocol logs that an attacker on a host cannot edit.

What Is Network Forensics?
Network forensics is the capture, storage, and analysis of traffic in order to reconstruct what happened during a security incident. Its central advantage is independence. An attacker with administrator rights can clear event logs, timestomp files, and uninstall the endpoint agent, but cannot retrospectively alter a packet capture held on a sensor they never reached.
In most investigations the network record answers three questions faster than anything else: which hosts spoke to the attacker’s infrastructure, how much data left the estate, and when the first contact occurred.
Two models exist, and mature teams use both.
Full packet capture records every byte. It permits file carving, credential recovery, and exact reconstruction of a session, but the storage arithmetic is unforgiving: a sustained 1 Gbps link generates roughly 10 TB per day. Few organisations retain full capture for more than a few days, and the usual compromise is to keep it only at the internet boundary and in front of the most sensitive segments.
Metadata is the practical alternative. Zeek produces structured logs for connections, DNS, HTTP, TLS, SMB, Kerberos, and files at perhaps one to two per cent of the volume of the raw traffic, while NetFlow or IPFIX from routers gives a thinner but far cheaper record. Six months of Zeek logs is more useful in a real investigation than three days of packets, because intrusions are typically discovered long after they begin.
The majority of traffic is now encrypted, and TLS 1.3 with Encrypted Client Hello removes the server name that used to make triage easy. What remains visible is still substantial: who talked to whom, when, for how long, in which direction the bytes flowed, the certificate chain where it is still observable, and the shape of the conversation.
TLS client fingerprinting, through JA3 and the newer JA4 family, identifies the implementation making the connection. Commodity tooling built on particular HTTP libraries produces distinctive fingerprints that differ from the browsers on the same host, which is often enough to spot an implant without reading a single byte of payload. Beacon analysis works on timing alone: a connection to the same destination every sixty seconds with five per cent jitter and a near-constant response size is a command channel regardless of what it carries.
Data exfiltration is the clearest example. When an organisation must tell the ICO within 72 hours whether personal data left the network, host evidence often cannot say. A flow record showing 40 GB transferred to an unfamiliar hosting provider over four nights, or the absence of any such transfer, is the difference between a notification that describes the facts and one that admits ignorance.
Network records also establish lateral movement, by showing SMB and WMI connections between workstations that have no business talking to one another, and they expose the full victim set, because every compromised host eventually contacts the same infrastructure.
Monitoring your own network is lawful in the UK, but it processes personal data and is subject to the UK GDPR and to the ICO’s guidance on monitoring workers. That means a documented lawful basis, normally legitimate interests, a data protection impact assessment where monitoring is systematic, a defined retention period, access restricted to the investigation team, and transparency in staff-facing policy. Decrypting employee TLS traffic raises the bar further and should be scoped, justified, and excluded for categories such as healthcare and banking.
Practically, plan placement before you need it. Capture at the internet boundary, between user and server segments, and in front of the systems that would end up in a breach notification. Synchronise every sensor to the same time source, because a timeline built from clocks that disagree by minutes will not withstand scrutiny. Pair the network record with memory forensics on the hosts it implicates, and you can usually explain an intrusion end to end.
At Privia Security, we build that visibility into SOC Maturation Services so the evidence exists before the incident does.
Full packet capture is usually retained for days rather than months because of storage cost, with seven days a common target at the internet boundary. Protocol metadata from Zeek or flow records should be kept far longer, ideally twelve months, since intrusions are often discovered well after they start. Set the retention period explicitly in policy and justify it in your data protection documentation.
Yes, with different methods. Connection metadata, timing, byte counts, directionality, certificate details, and TLS client fingerprints such as JA3 and JA4 remain available. Beaconing is detected through regular intervals and consistent response sizes rather than payload content. Encryption removes the ability to read messages, not the ability to establish who communicated with whom and how much data moved.
Network detection and response is a live monitoring capability that raises alerts as traffic passes. Network forensics is the retrospective examination of stored traffic records to reconstruct an incident, often months later. The same sensors frequently serve both purposes, but the forensic use case demands durable, searchable retention and defensible handling that a real-time detection product does not necessarily provide.
If the monitoring is systematic and extensive, which network capture across an employee population normally is, a data protection impact assessment is expected under UK GDPR. Document the lawful basis, the categories of data captured, retention, who can access it, and the measures limiting intrusion, such as excluding sensitive destinations from TLS inspection. The ICO’s guidance on monitoring workers sets out the expectations.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.