ISC · Security and Confidentiality and Privacy · 11 practice questions
Volatile Evidence Lost on Reboot: Incident Response
Capture volatile evidence before rebooting, powering off, or rebuilding a compromised server. Below: one server incident with changes to the evidence-preservation sequence and the resulting investigative loss.
Try one first
Hint
Which forensic artifacts exist only in volatile memory, and what happens to those artifacts when a system loses power?
Answer C. Powering off a running system destroys volatile memory and active session state that are only available during live response; those artifacts (in‑memory processes, session data, encryption keys) are often required to reconstruct attacker actions and timelines. While disk artifacts typically survive and can be imaged after shutdown, the absence of RAM and live‑state data substantially impedes a complete investigation and attribution.
Why not A: Tempting because stopping the system may limit further damage, but wrong because containment does not recover volatile artifacts (in‑memory keys, active connections, ephemeral indicators) needed for full root‑cause and timeline analysis; offline imaging alone can miss those critical pieces of evidence.
Why not B: Tempting because shutdown stops ongoing disk writes and some systems can produce hibernation or crash‑dump files, but wrong because an immediate power‑off does not reliably create those artifacts; volatile memory and active sessions are typically lost unless a controlled live capture, a deliberate hibernate, or a configured crash dump occurs.
Why not D: Tempting because some logs are flushed to disk at shutdown, but wrong because many important runtime details (in‑memory process state, ephemeral network sessions, packet captures) are not preserved and disk logs may be incomplete or overwritten; live collection captures a more complete evidentiary set.
Same scenario, one fact changes
Base case
Ridgeview Co. detects suspicious outbound transfers and a running process on its payroll server. Responders isolate the server and keep it powered on. No evidence has been collected, and live capture is feasible. They capture RAM, available process and connection records, host logs, and a forensic disk image before shutdown.
Answer: Both the captured volatile evidence and the imaged disk evidence are preserved.
Isolation limits ongoing harm. Live capture records evidence that shutdown would destroy, while the disk image preserves the host's disk contents for investigation.
Before you open each one, predict the answer.
Change 1Responders omit only the forensic disk image before shutdown.
Answer: Captured volatile evidence and collected records are preserved, but no original-host forensic disk image exists.
RAM capture and collected records remain usable. Ordinary shutdown generally leaves disk artifacts available for later imaging, but no disk image was created.
Change 2Responders reboot before collecting any evidence instead of collecting evidence before shutdown.
Answer: Uncaptured volatile evidence is lost; remaining disk artifacts may still support investigation.
Rebooting destroys live process and connection state, including memory-only credentials or keys. Disk evidence cannot fully replace those artifacts.
Change 3Responders immediately power off before collecting any evidence instead of collecting evidence before shutdown.
Answer: Uncaptured volatile evidence is lost, although disk artifacts generally survive for later imaging.
Power-off does not automatically save RAM or active sessions. Preserving the disk later cannot recreate activity that existed only in memory.
Change 4Responders wipe and rebuild before collecting any evidence instead of collecting evidence before shutdown.
Answer: Volatile evidence is lost, and original host disk artifacts are overwritten or altered.
Rebuilding replaces the compromised state. A later image captures the rebuilt system, not the original evidence needed for root-cause and scope analysis.
Key points
- Remember: A disk image preserves disk contents, not the live RAM and process state lost on reboot.
- Network isolation limits ongoing harm without requiring a shutdown.
- Capturing RAM does not create a forensic disk image.
- After a premature shutdown, preserve remaining disk evidence and logs rather than assuming everything is lost.
How the exam traps you
- Assuming a later disk image replaces missing RAM. Disk evidence generally survives shutdown, but it cannot replace uncaptured memory-only artifacts.
- Rebooting first because stopping the suspicious process preserves evidence. Stopping a process is not capturing it. Preserve its live state before rebooting when feasible.
- Treating a rebuilt server as the original forensic evidence. A post-rebuild image records the rebuilt state, not the original compromised state.
Now the same facts as questions
Each question changes one fact from the one before. Watch which change flips the answer.
Question 1
Answer B. Live capture preserves volatile artifacts, and the forensic disk image preserves disk contents.
Why not A: Shutdown destroys live state, not an already saved RAM capture.
Why not C: These captures support investigation but do not guarantee a complete account of attacker activity.
Why not D: Shutdown does not invalidate an already preserved forensic disk image.
Question 2
Answer A. The listed captures exist, but responders did not create a forensic disk image.
Why not B: A RAM capture does not create an image of disk contents.
Why not C: Captured RAM and collected records remain usable despite the missing disk image.
Why not D: Ordinary shutdown generally leaves disk artifacts available for later imaging.
Question 3
Answer C. Rebooting destroys uncaptured live state without necessarily destroying persistent disk evidence.
Why not A: A reboot does not reliably create a capture of the original RAM and connection state.
Why not B: Stopping a process does not establish what it did or what data it accessed.
Why not D: Disk artifacts cannot fully replace evidence that existed only in memory.
Question 4
Answer D. Power-off destroys uncaptured volatile state, but persistent disk evidence generally remains.
Why not A: Power-off does not automatically capture RAM, processes, or active connections.
Why not B: Ordinary power-off generally does not erase persistent disk contents.
Why not C: A disk image cannot recreate every session that existed only in memory.
Question 5
Answer B. Rebuilding eliminates live state and replaces or changes original host artifacts.
Why not A: A clean operational state is not a preserved copy of the compromised state.
Why not C: Wiping and rebuilding directly changes disk contents, including potential host evidence.
Why not D: A post-rebuild image captures the replacement state rather than the original compromised host.
Question 2
Hint
Which immediate action best preserves the ability to reconstruct the incident and support defensible reporting?
Answer D. The immediate priority is preserving and securing whatever forensic evidence remains and documenting chain-of-custody. Disabling accounts and rebooting the server without capturing volatile memory, imaging disks, or collecting logs likely destroyed or compromised critical evidence; securing what is left is necessary to determine scope, support root-cause analysis, and provide defensible information for regulators and legal processes. Notification, vendor actions, and system recovery should follow containment and evidence preservation.
Why not A: This is tempting because breach notifications are time-sensitive and regulators/individuals expect prompt disclosure. However, preparing or sending notifications before preserving evidence can result in inaccurate scope reporting and may hinder investigation and legal defensibility; the team must secure remaining evidence first so reporting is based on reliable findings.
Why not B: This seems sensible as a containment and accountability step. But access was already disabled and terminating the vendor now does not recover lost volatile evidence; abrupt contractual actions can also interfere with evidence collection if the vendor holds logs or other data the investigation needs. Securing evidence takes precedence.
Why not C: Restoring systems quickly is attractive to reduce downtime, but recovery before evidence preservation risks overwriting remaining forensic data and prevents accurate root-cause analysis. Recovery should occur only after containment and careful evidence collection.
Question 3
Hint
Separate the value of containment actions from the value of evidence preservation. Ask which evidence source cannot be recreated after the sequence described.
Answer C. Isolating the server and resetting credentials can be appropriate containment actions, but reimaging before collecting host evidence sacrifices important forensic data. Memory contents, live processes, transient connections, and certain host artifacts may be unrecoverable after the rebuild. Even with good centralized logs, the team may be less able to reconstruct exactly what the attacker did on that host or how persistence was established.
Why not A: This distractor is attractive because credential reset is commonly viewed as a containment step rather than an evidence issue. But the larger problem is the reimage before preservation, which can materially impair root-cause analysis and detailed scoping of host activity. Complete network logs still do not capture all local persistence, execution, and in-memory evidence.
Why not B: This is tempting because candidates often over-rely on centralized logging and assume isolation 'freezes' the evidence picture. However, a post-reimage image captures the rebuilt state, not the compromised host state. Central logs help with scoping, but they do not fully replace lost memory and host-resident forensic artifacts.
Why not D: Candidates may confuse evidence preservation mistakes with total chain-of-custody failure. Reimaging the host certainly harms the evidentiary record for that system, but it does not automatically make separately preserved centralized logs unusable. Those logs can still support investigation if they were retained and handled appropriately.
Question 4
Hint
Think about how to stop ongoing data loss quickly while preserving volatile evidence for investigation.
Answer A. Isolating the server stops further data exfiltration while keeping it powered preserves volatile evidence (RAM, active connections, and recent logs) needed for a proper forensic investigation and attribution. This containment approach balances halting the incident with preserving the chain of custody for investigators. Other immediate actions risk destroying evidence or spreading premature information.
Why not B: Powering off may halt traffic, but it destroys volatile data (memory contents, active network state) that are critical to understanding the attack and attributing it. Investigators generally prefer isolation while the system remains powered unless there is an overriding safety reason to shut down.
Why not C: Notifications are important but premature as the first step; they should follow containment and an initial assessment of scope to avoid spreading inaccurate information and to meet legal requirements with verified facts.
Why not D: A firewall rule might help but changing live configurations can alter evidence and may not completely stop the attacker. Proper incident response typically begins with isolating the affected system under forensic guidance; network controls can be part of containment after preservation steps are taken.
Question 5
Hint
Which forensic artifacts exist only in RAM or in active network sessions, and how does losing those artifacts limit investigators' ability to reconstruct actions or recover decrypted material?
Answer D. A reboot clears RAM and other volatile state, destroying artifacts such as running processes' memory, decrypted material, session tokens, and live network connections that exist only in memory. Those in‑memory artifacts are often essential to reconstruct an attacker's actions, establish timing, and link activity to exfiltration or attribution. Persistent storage and logs can help but typically cannot replace the transient, in‑memory evidence lost on reboot.
Why not A: Tempting because centralized logs and agents can provide valuable context, but wrong to assume they fully compensate: logs and agents often lack ephemeral memory contents, may be incomplete or not cover all activity, and cannot reconstruct decrypted in‑RAM artifacts or live process state.
Why not B: Tempting because disk images and backups are reliable sources of persistent evidence and support recovery, but wrong: they do not capture volatile, in‑memory state (e.g., decrypted content, process memory, ephemeral connections) that may be required to understand what the attacker did in RAM or to prove exfiltration.
Why not C: This is plausible since a reboot can stop running malware and reduce symptoms, but it is misleading as the primary consequence: reboot destroys volatile evidence investigators need to confirm eradication or identify persistence mechanisms, and many attacks include mechanisms that survive a reboot.
Question 6
Hint
Focus on incident-response sequencing when active harm is ongoing and some evidence will disappear if you move too quickly to recovery.
Answer A. Active data exfiltration is occurring, so the immediate priorities are to stop further loss while preserving perishable forensic data. Isolating the affected host limits ongoing harm, while capturing volatile memory, preserving cloud session tokens, and saving logs that will roll off preserves critical evidence needed to scope the breach and identify persistence. Recovery actions such as rebuilding from backup should follow containment and evidence preservation.
Why not B: Tempting because it appears to balance continuity with a limited response, but disabling a single account may not stop an active compromise and leaves a compromised system online during ongoing exfiltration. Partial measures that fail to contain the host risk continued data loss and hamper accurate scoping of the incident.
Why not C: Tempting under operations pressure and given a clean backup, but rebuilding immediately destroys volatile evidence (memory, live session data) and may prevent saving short-retention logs needed to determine entry points and persistence. Recovery should be deferred until necessary forensic data are captured and containment is achieved.
Why not D: Tempting because SSNs are highly sensitive, but the stem specifies that external notifications are not yet required until scope is reasonably assessed. Addressing the active exfiltration and preserving evidence to define scope comes first; notifications may follow once the extent of exposure and legal obligations are clear.
Common questions
What volatile evidence is lost when a server reboots?
Uncaptured RAM contents and live state can be lost, including running-process state, active connections, in-memory credentials, and ephemeral session keys. Previously collected evidence remains available.
Does shutting down a server destroy its disk evidence?
Disk artifacts generally survive an ordinary shutdown and can be imaged later. That later image does not recover live RAM or connection state that was never captured.
Why capture evidence before reimaging a compromised server?
Reimaging can overwrite host logs and other original disk artifacts while also eliminating volatile state. A rebuilt server cannot substitute for captures of the compromised host.
Practice ISC like the real exam
The free ChatCPA simulator: real exam layout, timed testlets, starting with a question on this topic. No account needed to start.
Open the free simulator →More on Incident response
Questions from the ChatCPA bank of 17,658 CPA exam questions, each with a written reason for every wrong answer. ChatCPA is built by Nicholas Miller, CPA (Oregon #14907). How these pages are made. Spot an error? Tell us.