Showing raw Markdown: the renderer could not load.
## Questions
### 1. A colleague says: "NIST replaced the whole incident response life cycle with the CSF, so the four phases of the old guide are wrong and useless." Which reply is most accurate?
*Objective: 1. Explain the incident response life cycle in NIST SP 800-61r3 and how it relates to the NIST Cybersecurity Framework 2.0 functions.*
- A) Correct: SP 800-61r3 deletes preparation and post-incident activity.
- B) Incorrect: SP 800-61r3 is only a minor edit that adds a "Govern" phase between detection and containment.
- C) Partly: SP 800-61r3 (2025) supersedes r2 and organises incident response around the CSF 2.0 functions, but the four classic phases still describe well what teams do hour by hour.
- D) Incorrect: SP 800-61r3 applies only to cloud providers, so r2 is still the standard for everyone else.
### 2. After a phishing incident, the team rewrites its playbook, adds mail filtering rules and updates the list of critical assets. In the CSF 2.0 frame, where do these improvements mainly belong?
*Objective: 1. Explain the incident response life cycle in NIST SP 800-61r3 and how it relates to the NIST Cybersecurity Framework 2.0 functions.*
- A) They feed back into Govern, Identify and Protect, the "getting ready" side, so the next incident is handled better.
- B) They belong to Detect, because they find new attacks.
- C) They belong to Recover, because the incident is over.
- D) They are not part of any function: improvement is outside incident response.
### 3. A ransomware attack encrypts the file server. Staff can't work for a full day. The backups are intact, and restoring them needs two extra technicians from a contractor, but the time is predictable. No data left the company. Which classification (functional impact / information impact / recoverability) fits best, using the SP 800-61r2 categories?
*Objective: 2. Classify and prioritise an incident by its functional impact, the data affected and the effort needed to recover.*
- A) Low / None / Regular
- B) None / Privacy breach / Not recoverable
- C) High / Proprietary breach / Extended
- D) High / None / Supplemented
### 4. On Monday at 09:00 you confirm that an attacker downloaded a customer list with names and email addresses from your company's web application. The application is running normally. You operate in the EU. Which is the best next step?
*Objective: 2. Classify and prioritise an incident by its functional impact, the data affected and the effort needed to recover.*
- A) Treat it as low priority: nothing is down, so functional impact is none.
- B) Treat it as high priority, and involve the data protection officer or legal contact at once, because a personal data breach may have to be notified to the supervisory authority within 72 hours of becoming aware of it.
- C) Wait until the full forensic report is finished before telling anyone, so you don't pass on wrong information.
- D) Delete the customer list from the web application so that it can't be stolen again.
### 5. You find these three lines for the same address. What is the most reasonable conclusion?
*Objective: 3. Analyse authentication and web server logs to build a timeline of an incident.*
```text
Nov 4 02:11:08 web01 sshd[2312]: Failed password for alice from 203.0.113.45 port 41526 ssh2
Nov 4 02:11:09 web01 sshd[2341]: Failed password for bob from 203.0.113.45 port 49489 ssh2
Nov 4 02:14:28 web01 sshd[2926]: Accepted password for dave from 203.0.113.45 port 50280 ssh2
```
- A) A user who forgot three different passwords, then remembered one.
- B) A brute-force attack against `alice` only, which failed.
- C) Probable password spraying across several accounts, ending in a compromise of `dave`, which needs urgent verification and containment.
- D) An error in the log: failed and accepted logins can't come from the same address.
### 6. In the pipeline `grep 'Failed password' auth.log | grep -oE 'from [0-9.]+' | awk '{print $2}' | sort | uniq -c | sort -rn`, what would go wrong if the plain `sort` before `uniq -c` were removed?
*Objective: 3. Analyse authentication and web server logs to build a timeline of an incident.*
- A) `uniq -c` only counts identical lines that are next to each other, so repeated addresses scattered in the file would be counted as separate groups and the totals would be wrong.
- B) Nothing: `sort -rn` at the end would fix the order.
- C) `awk` would stop printing the addresses.
- D) `grep -oE` would print the whole line instead of the address.
### 7. A server you believe is compromised is still running. You decide to collect evidence before changing anything. Which order follows RFC 3227?
*Objective: 4. Choose containment, eradication and recovery actions, and explain how to preserve evidence while doing so.*
- A) Disk image, then memory, then network connections, then remote logs.
- B) Network connections and running processes, then memory, then a disk image, then logs held on a remote server.
- C) Remote logs, then a disk image, then memory, then network connections.
- D) Power the server off, then take a disk image, then collect remote logs.
### 8. Your analysis shows that an attacker is copying data from a database server right now. The business can tolerate an outage of a few hours. What is the best approach?
*Objective: 4. Choose containment, eradication and recovery actions, and explain how to preserve evidence while doing so.*
- A) Monitor quietly for a few days to learn who the attacker is before acting.
- B) Reinstall the server immediately to remove the attacker, then investigate using the new installation.
- C) Block the attacker's IP address and consider the incident solved.
- D) Record the volatile state and copy the logs quickly, then isolate the server from the network (not power it off), and plan eradication and recovery afterwards.
### 9. In a blameless post-incident review, an engineer says: "I clicked a link in a phishing email because I was in a hurry." Which response best fits the blameless approach?
*Objective: 5. Write a short incident report with root cause, impact and lessons learned.*
- A) Record the engineer's name as the root cause, so that everyone knows who to train.
- B) Ask the engineer to leave the meeting so the others can speak freely.
- C) Ask what in the system made the click harmful, for example no link protection, no second factor and no quick way to report, and turn the answers into owned actions.
- D) Skip the review, because human error can't be fixed.
### 10. Which entry in an incident report is a **root cause** rather than a symptom or an effect?
*Objective: 5. Write a short incident report with root cause, impact and lessons learned.*
- A) "4.8 MB of customer data was downloaded."
- B) "The admin pages and SSH accepted a password alone, with no second factor and no lockout after repeated failures."
- C) "The attacker logged in at 02:14."
- D) "The incident was detected at 08:00 the next morning."
---
## Answer key
<details>
<summary>Show answers</summary>
| # | Answer | Explanation |
|---|---|---|
| 1 | C | SP 800-61r3 (April 2025) supersedes r2 and frames incident response with the CSF 2.0 functions, but preparation, detection, containment and lessons learned are still what teams do, so the four phases remain a useful mental model. A and D are false; B invents a "Govern phase". |
| 2 | A | Lessons learned improve the preparation side: Govern (policy), Identify (assets, improvements) and Protect (safeguards). Detect, Respond and Recover are what you do during the incident. |
| 3 | D | A full day of lost work is high functional impact, no data left (information impact none), and recovery is predictable but needs extra resources, which is "Supplemented". "Regular" would mean no extra resources; "Extended" means unpredictable time. |
| 4 | B | Stolen personal data is a privacy breach and the loss can't be undone, so priority is high even with no downtime. Under the GDPR, the 72-hour notification clock runs from becoming aware, and whether it applies is a legal judgement, so involve the DPO or legal early. Deleting data (D) destroys evidence and doesn't undo the theft. |
| 5 | C | One address tries different accounts within seconds and then logs in as another account: the pattern of password spraying followed by a successful guess. It needs verification, but it is no coincidence. |
| 6 | A | `uniq -c` collapses only adjacent duplicates. Without `sort`, the same address appearing in different places would be counted several times. The final `sort -rn` only orders the groups, it doesn't merge them. |
| 7 | B | RFC 3227 orders from most to least volatile: connections, processes and memory first, then temporary files, disk, and remote logs. A reverses it; D loses memory and state by powering off first. |
| 8 | D | Active data loss means waiting is costly, so contain now: but record the volatile state first and keep the machine powered on (isolate it from the network rather than turning it off). Blocking one IP (C) is easy to bypass, and a reinstall (B) destroys the evidence. |
| 9 | C | The blameless approach asks what in the system allowed a normal human action to cause harm, and fixes that. Naming and blaming (A) teaches people to hide mistakes. |
| 10 | B | A root cause is a condition you can fix that made the incident possible. A is impact, C is a timeline event and D describes detection delay, which is a contributing factor, not the cause. |
</details>