Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is Memory Forensics?

Memory forensics is the examination of a captured RAM image to reconstruct what a system was doing: hidden processes, injected code, network state, and secrets that never reach disk.

Illustration for the cybersecurity term Memory Forensics
Home / Blog / Cybersecurity Terms / What Is Memory Forensics?
17 November 2026

What Is Memory Forensics?

What Is Memory Forensics?

Memory forensics is the analysis of a captured RAM image to work out what a computer was actually doing at the moment of capture. Disk tells you what was stored; memory tells you what was running. The distinction decides most intrusion investigations, because a great deal of current tradecraft is designed specifically to avoid writing anything useful to the file system.

A PowerShell loader that downloads an encrypted blob, decrypts it in memory, and executes it without ever creating a file leaves almost nothing for a disk examiner. The same activity leaves an unmistakable trail in RAM: an executable region with no backing file, the decrypted configuration in plain text, and a socket to the command server.

What an Image Gives You

The Toolset

Volatility 3 is the open-source standard. Unlike its predecessor it does not need a hand-picked profile; it resolves kernel structures from symbol tables generated from the operating system’s own debug symbols, downloading the matching Windows ISF file automatically. Linux images still need a symbol table built from the exact kernel with dwarf2json, which is a common source of frustration when the kernel version was not recorded at capture time.

MemProcFS takes a different approach, mounting the image as a virtual file system so analysts can browse processes and handles with ordinary tools, and Velociraptor runs comparable logic on live endpoints at scale. Commercial suites such as Magnet AXIOM and Belkasoft X wrap similar engines in a review interface.

A Practical Order of Work

  1. Confirm the image parses and identify the build with windows.info.
  2. List processes two ways. windows.pslist walks the active process list; windows.psscan scans pool allocations and finds entries that have been hidden or have exited. A process present in one and absent from the other is worth explaining.
  3. Review windows.cmdline for encoded PowerShell, unusual parent-child pairs, and binaries running from user-writable directories.
  4. Hunt for injection with windows.malfind, then verify each hit rather than trusting it.
  5. Pull network state with windows.netscan and pivot on any address that does not belong to the estate.
  6. Dump suspicious regions and processes to disk for static analysis and hashing.

Where Memory Analysis Earns Its Place

It is the most reliable way to deal with fileless attacks, packed or obfuscated samples that only reveal themselves once unpacked, and kernel-level rootkits that lie to every tool running on the live system. It also recovers the configuration blocks of commodity frameworks: a Cobalt Strike beacon’s sleep time, jitter, and command server list sit in memory in a parsable structure, which turns one infected host into a network-wide hunting signature.

Memory also answers questions that logs cannot. When an organisation needs to prove whether an attacker reached a particular database, the connection object, the query string, and the returned rows may all still be resident.

Known Limits

An image is a snapshot, not a recording. Anything paged out, overwritten, or executed before capture may be absent, and there is no way to tell the difference between never present and already gone. Timestamps are scarce: a closed socket recovered by pool scanning carries no reliable time, so findings must be corroborated against timeline analysis of disk and log sources.

Smear during acquisition can corrupt individual structures, and anti-forensic techniques exist that deliberately damage the structures scanners rely on. Treat memory as one strong source among several rather than a final authority.

Our analysts use memory evidence alongside endpoint and network telemetry during Investigation and Incident Response Services, because the three together close gaps that none of them closes alone.

FAQ – Frequently Asked Questions

What Is the Difference Between Memory Forensics and Disk Forensics?

Disk forensics recovers what was stored: files, deleted data, file system metadata, and registry hives. Memory forensics recovers what was executing: processes, injected code, open connections, and decrypted secrets. Many intrusions leave almost nothing on disk by design, so the two disciplines answer different questions and a serious investigation normally uses both against the same host.

Do I Need Volatility Profiles for Windows Images?

Not with Volatility 3. It builds structure definitions from Intermediate Symbol Format files derived from Microsoft debug symbols, and it downloads the one matching your image automatically. Linux and macOS are different: you still need a symbol table generated from the exact kernel that was running, so record the kernel version during acquisition or you may be unable to analyse the image at all.

Can Memory Forensics Recover Encryption Keys?

Often, yes. Full disk encryption keys, TLS session keys, and SSH keys must exist in plaintext in RAM for the system to function, so they can be extracted from an image taken while the volume was mounted or the session was open. This is one reason an investigator captures memory before shutting a machine down, and also why cold boot attacks remain a credible physical threat.

How Quickly Does Memory Evidence Decay?

Continuously. Terminated process structures survive in pool memory for seconds to hours depending on system load, network connection objects are reused quickly on busy servers, and paging moves data to disk unpredictably. A capture taken an hour after detection will hold meaningfully less than one taken immediately, which is why acquisition is treated as an urgent task rather than a scheduled one.

Related Terms

You May Be Interested In These

What Is ISO/IEC 27005?

Cybersecurity Terms

17 November 2026

What Is ISO/IEC 27005?

Read More
What Are Amcache and ShimCache?

Cybersecurity Terms

17 November 2026

What Are Amcache and ShimCache?

Read More
What Happened in the Change Healthcare Attack?

Cybersecurity Terms

17 November 2026

What Happened in the Change Healthcare Attack?

Read More