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

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is Volatile Data Collection?

Volatile data collection gathers the evidence that disappears when a system is powered down, in a defined order, with every command recorded so the method can be defended later.

Illustration for the cybersecurity term Volatile Data Collection
Home / Blog / Cybersecurity Terms / What Is Volatile Data Collection?
17 November 2026

What Is Volatile Data Collection?

What Is Volatile Data Collection?

Volatile data is anything that exists only while a system is running. Memory contents, running processes, open network connections, logged-on users, mounted shares, cached credentials, clipboard contents, and the routing table all vanish the moment the power goes. Volatile data collection is the disciplined gathering of that material before it is lost.

It is the point at which incident response differs most sharply from traditional computer forensics. The classical advice was to pull the plug and image the disk. Applied to a modern intrusion, that advice destroys the majority of the useful evidence.

The Order of Volatility

RFC 3227, published in 2002 and still the cited reference, sets out the sequence by how quickly each source decays:

  1. CPU registers, cache, and the contents of memory.
  2. Routing tables, ARP cache, process table, kernel statistics, and active connections.
  3. Temporary file systems.
  4. Data on disk.
  5. Remote logging and monitoring data relevant to the system.
  6. Physical configuration and network topology.
  7. Archival media.

In practice this collapses into a simple rule for the field: take the memory image first, because it captures nearly everything in categories one and two; then run a short, scripted set of live commands to record system state in a readable form; then move to disk.

What to Capture Beyond Memory

A memory image needs parsing before it yields answers, so a live state capture is still worth the few minutes it takes. On Windows the useful set includes the process list with full command lines and parent relationships, network connections with owning processes, logged-on sessions, local and domain group membership of active accounts, services and scheduled tasks, installed drivers, autoruns, open file handles, SMB sessions and shares, the ARP cache, and the DNS resolver cache.

On Linux the equivalents are the process tree with command lines, ss output for sockets, loaded kernel modules, mounted file systems, cron entries, systemd units, logged-in sessions, and the contents of /proc for any suspicious process, including its executable link and environment.

Cloud and container workloads add their own volatile layer: running container inventories, the instance metadata service response, and temporary role credentials that will expire within hours.

Doing It Without Contaminating the Evidence

Deciding Between Containment and Collection

The tension in every live incident is that isolating a host protects the business while collecting from it preserves the evidence. The resolution is usually to isolate at the network layer first, using an EDR containment function or a switch port change, which stops the attacker interacting with the host while leaving it running, and then collect.

Where active destruction is under way, containment wins without argument. Where an attacker has an interactive session, assume they may notice the collection and act accordingly.

Prepare It in Advance

The common failure is improvisation: a first responder working out which commands to run while a director asks for updates every ten minutes. Build the collection script, test it on a representative build, store it on prepared media or deploy it through the EDR platform, and rehearse it in a tabletop exercise. Agree in advance who has authority to run it on a production system at three in the morning.

Done properly, this half hour of work is frequently what allows the subsequent memory and disk analysis to produce a definite answer rather than a range of possibilities, which is why it is the first operational step in our Investigation and Incident Response Services.

FAQ – Frequently Asked Questions

What Counts as Volatile Data?

Anything that exists only while the system is powered on: memory contents, running processes and their command lines, open network connections, logged-on sessions, cached credentials, clipboard contents, the ARP and DNS caches, mounted shares, and temporary role credentials in cloud environments. All of it is lost on shutdown, and much of it changes significantly within minutes even while the system stays running.

What Is the Order of Volatility?

The sequence set out in RFC 3227 for collecting evidence from most to least perishable: registers and cache, then memory, then network and process state, then temporary file systems, then disk, then remote logs, then archival media. Following it means capturing memory before running live commands, and running live commands before imaging the disk.

Does Running Collection Tools Damage the Evidence?

It changes the system, which is unavoidable and accepted. Executing any binary allocates memory, creates process artefacts, and may update file access times. The requirement is not to leave the system untouched but to document precisely what was run, when, and by whom, so the changes can be explained and distinguished from the attacker’s activity.

Should We Isolate a Host Before or After Collecting Volatile Data?

Isolate first at the network layer where possible, using endpoint detection containment or a switch configuration change, then collect. This stops the attacker interacting with the host while leaving it running so the evidence survives. Full shutdown or disconnection of power should be reserved for cases where destruction is actively occurring and evidence loss is the lesser harm.

Related Terms

You May Be Interested In These

What Is iOS Forensic Analysis?

Cybersecurity Terms

17 November 2026

What Is iOS Forensic Analysis?

Read More
What Are Black Start and Grid Resilience?

Cybersecurity Terms

17 November 2026

What Are Black Start and Grid Resilience?

Read More
What Is Firmware Security?

Cybersecurity Terms

17 November 2026

What Is Firmware Security?

Read More