Vulnerability Management: CVE, CVSS, Scanning and Patching
Every organisation runs software with flaws. New vulnerabilities are published every day, and attackers move quickly to use the ones that matter. Vulnerability management is the continuous process of finding weaknesses, deciding which ones matter most and fixing them in time. It is one of the highest-value activities in security, because most breaches involve a weakness that was already known and fixable. This lesson explains the vocabulary, the scoring systems and the process, with practice you can do in your lab.
Key Terms
| Term | Meaning |
|---|---|
| Vulnerability | A weakness in software, hardware, configuration or process |
| CVE (Common Vulnerabilities and Exposures) | A unique public ID for a known vulnerability, such as CVE-2021-44228 |
| CWE (Common Weakness Enumeration) | A catalogue of weakness types (for example CWE-89, SQL injection) |
| CVSS (Common Vulnerability Scoring System) | A 0 to 10 score of technical severity |
| NVD (National Vulnerability Database) | NIST’s database enriching CVEs with scores and affected products |
| Advisory / patch / hotfix | Vendor announcement and fix |
| Zero-day | A vulnerability exploited before a fix is available |
| N-day | A known vulnerability that has a fix but remains unpatched |
| Exposure | A vulnerability that is actually reachable by an attacker |
Understanding CVSS
CVSS provides a standard way to describe severity. The base score reflects intrinsic characteristics: how the flaw is reached (network, local), complexity, privileges and user interaction needed, and impact on confidentiality, integrity and availability. Severity bands:
| Score | Rating |
|---|---|
| 0.0 | None |
| 0.1 – 3.9 | Low |
| 4.0 – 6.9 | Medium |
| 7.0 – 8.9 | High |
| 9.0 – 10.0 | Critical |
Versions 3.1 and 4.0 of CVSS are both in use. A CVSS vector string such as CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H reads: network-reachable, low complexity, no privileges, no user interaction, scope unchanged, high impact on all three properties. That combination scores 9.8. The limitation is that CVSS measures severity in general, not risk in your environment. A critical flaw in a system that is isolated and unused may matter less than a medium flaw on your public login server.
Risk-Based Prioritisation
Thousands of vulnerabilities cannot all be fixed immediately. Smart teams combine multiple signals:
- CVSS severity: how bad could it be?
- Exploitation evidence: the CISA Known Exploited Vulnerabilities (KEV) catalogue lists flaws seen exploited in real attacks; these should rise to the top.
- EPSS (Exploit Prediction Scoring System): a probability that a vulnerability will be exploited in the near future.
- Exposure: is the system internet-facing, or reachable from low-trust networks?
- Asset criticality: what data or business function does it support?
- Compensating controls: is it protected by segmentation, a WAF or disabled features?
- Exploit availability: are public tools or proof-of-concepts known?
A simple scoring approach you can adopt:
| Priority | Typical criteria | Example target time |
|---|---|---|
| P1 Emergency | In the KEV catalogue or actively exploited, internet-facing or critical asset | Hours to a few days |
| P2 High | Critical or high severity, reachable, exploit likely | 7 to 14 days |
| P3 Medium | Medium severity, or high but well controlled | 30 days |
| P4 Low | Low severity or very limited exposure | 90 days or next maintenance cycle |
Your own policy should define target timelines and approval for exceptions.
The Vulnerability Management Life Cycle
- Discover assets: you cannot scan what you do not know about. Maintain an inventory of systems, software and cloud resources.
- Scan and assess: run authenticated and unauthenticated scans on a regular schedule, plus scanning of containers, code dependencies and cloud configuration.
- Validate: remove false positives and confirm that findings are real.
- Prioritise: apply the risk-based method above.
- Remediate: patch, reconfigure, upgrade, or apply a mitigation or workaround.
- Verify: rescan to confirm the fix worked.
- Report and improve: track metrics and fix root causes, such as poor patching processes.
Scanner Types and Tools
| Scan type | What it covers | Examples |
|---|---|---|
| Network vulnerability scanning | Hosts, services, missing patches, misconfigurations | OpenVAS / Greenbone (free), Nessus, Qualys, Rapid7 |
| Web application scanning (DAST) | Running web apps | OWASP ZAP, Burp Suite, commercial scanners |
| Software composition analysis (SCA) | Vulnerable open-source libraries | OWASP Dependency-Check, Trivy, Dependabot, Snyk |
| Container and image scanning | Vulnerabilities in images | Trivy, Grype |
| Cloud posture management | Misconfigurations in cloud accounts | Provider-native tools, Prowler, ScoutSuite |
Authenticated scans log in to the system and give far more accurate results about missing patches than unauthenticated scans, which see only what an outsider sees. Use both. Scanners can disrupt fragile systems, so schedule scans and test in staging first.
Practice Lab 1: Look Up a Real Vulnerability
- Open the National Vulnerability Database (nvd.nist.gov) and search for
CVE-2021-44228(the widely documented Log4Shell vulnerability). - Record its CVSS score, vector and affected products. Read the CWE listed.
- Open the CISA KEV catalogue (cisa.gov/known-exploited-vulnerabilities-catalog). Is it listed? What does the catalogue suggest about prioritisation?
- Write a short prioritisation note: if you ran 50 servers, how would you find out which are affected, and in what order would you act?
Practice Lab 2: Scan Your Own Lab Target
Run a vulnerability scan only against your own lab VMs. A quick, free route is Greenbone/OpenVAS installed on your analyst or monitoring VM, or a command-line scan with Nmap’s vuln category scripts, which look up known vulnerabilities by service version. Keep your target on the host-only network:
nmap -sV --script vuln 192.168.56.102 -oA lab-vuln-scan
Review the results and fill in a tracking table:
| Finding | Affected asset | CVSS | Exposure | Priority | Fix | Status |
|---|---|---|---|---|---|---|
| Outdated web server version | lab-web-01 | 7.5 | Reachable by all lab hosts | P2 | Upgrade to supported version | Open |
| Cleartext management service enabled | lab-web-01 | 6.5 | Reachable by all lab hosts | P3 | Disable service, use SSH | Open |
Then fix one item, rescan and confirm the finding disappears. That remediate-and-verify loop is the real skill.
Patch Management Essentials
- Define patch windows and emergency procedures for critical flaws.
- Test patches in a staging environment that mirrors production where practical.
- Have a rollback plan, and use formal change control so that every patch is approved, scheduled and recorded.
- Prioritise internet-facing systems, remote-access devices and widely exploited software.
- Do not forget firmware, network devices, third-party applications, browsers and libraries.
- When patching is impossible, apply compensating controls: isolate, restrict access, disable features, add monitoring.
- Remove end-of-life software, since it will never be fixed.
Metrics That Matter
- Mean time to remediate by priority level.
- Percentage of assets scanned and scan coverage.
- Number of open critical and high findings and their age.
- Percentage of KEV vulnerabilities remediated within target.
- Recurrence rate of the same issue types, which signals process problems.
Common Mistakes
- Scanning without an asset inventory, so whole systems are missed.
- Treating CVSS as the only factor, so low-score but actively exploited flaws are ignored.
- Drowning teams in thousands of unvalidated findings.
- Reporting without verification rescans.
- Leaving unsupported software running because “it still works”.
Practice Exercises
- Decode a CVSS vector: what does
AV:L/AC:H/PR:H/UI:Rsuggest about how easy it is to attack? - Given two findings, (a) critical, no known exploitation, on an isolated test server, and (b) medium, in the KEV catalogue, on your public website, which do you fix first, and why?
- Write a one-page vulnerability management policy covering scanning frequency, priority levels and exception handling.
- Compare an authenticated and an unauthenticated scan of the same lab VM. What does each miss?
Mentor tip: Most organisations do not fail because they lack a scanner; they fail because findings never become fixes. Build the relationship with system owners first, and make the fix easy.
Educational use only: scan only systems you own or are authorised to assess.
- CVE identifies a vulnerability, CWE its weakness type, and CVSS its technical severity.
- Prioritise using severity, exploitation evidence (KEV, EPSS), exposure and asset criticality.
- Run the cycle: discover, scan, validate, prioritise, remediate, verify and report.
- Authenticated scanning and good asset inventories improve accuracy.
- Findings only matter when they are fixed and verified.