Week 10 · Lesson 3

Upgrade, Update & Patch

Every piece of software ships with defects, and every defect is a potential opening for an attacker. The single most effective thing an organisation can do is keep software current: identify what is vulnerable, test the fix, and roll it out before someone else rolls out an exploit. This lesson defines the vocabulary precisely, walks the patch process, and shows the exact Cisco IOS commands to back up and upgrade a device.

CCNA 200-301 Domain 5.0 Security Fundamentals Verified 2026-10-11
Part 1

Vulnerability, exploit, threat — and why software becomes vulnerable

These three words are used loosely in everyday speech and precisely in security exams. Learn the canonical definitions now, because the whole discipline is built on them.

The vocabulary · select a term

Put them in one sentence: a threat uses an exploit to take advantage of a vulnerability. The three combine into risk, the likelihood and impact of loss. A common way to express it is risk = (threat × vulnerability) × impact: remove the vulnerability (patch it), and the risk from that threat drops to near zero even if the threat still exists. That is exactly why patching is the highest-value defensive activity available.

TermPlain meaningExample
ThreatWhat might cause harm (an actor or an event).A remote attacker, or a worm spreading on the Internet.
VulnerabilityThe weakness that makes harm possible.A buffer overflow in an old IOS release; a default password left in place.
ExploitThe concrete method that abuses the weakness.Crafted packets that trigger the overflow and grant remote code execution.
RiskThe chance and size of the loss if it happens.A high-CVSS flaw exposed to the Internet on an unpatched edge router.
Related vocabulary. A zero-day is an exploit used before the vendor releases a patch — defenders have zero days of warning. An n-day is an exploit that appears after a patch exists but before it is widely applied. N-days cause the majority of real breaches, precisely because organisations lag on patching.

Why software becomes vulnerable

Software is written by humans, and humans make mistakes. The main sources of weakness are:

Definition — end-of-life (EOL). When a vendor ends software maintenance, it stops producing bug fixes and security patches for that release. When a product reaches its Last Date of Support (LDoS), there are no further fixes at all. Software past these dates is EOL, and every vulnerability discovered after that point becomes a permanent one.

EOL is not the same as “broken”. An EOL device may still forward packets perfectly well, which is exactly why it lingers in production for years while slowly accumulating unpatched, publicly documented flaws. From a security standpoint, an EOL device is an unpatchable device, and an unpatchable device should be replaced on a managed schedule.

Part 2

The vulnerability and patch lifecycle

A vulnerability is not a single event; it moves through recognisable stages. The period between a fix being available and that fix being applied is the patch gap (also called the window of exposure) — the time an attacker can still win.

The vulnerability lifecycle · step through the stages
1 / 7

StageWhat happensDefender action
1. IntroductionThe defect ships in the code. Nobody knows it is there.Buy from supported vendors; track versions.
2. DiscoveryA researcher or attacker finds the flaw.Maintain asset inventory so you can react fast.
3. DisclosureA CVE is assigned and the vendor publishes a security advisory.Subscribe to Cisco PSIRT advisories and CVE feeds.
4. ExploitationAn exploit or proof of concept appears; attacks begin.Prioritise by CVSS score and exposure; apply mitigations.
5. Patch releaseThe vendor ships a fixed release or patch.Test in the lab; plan the maintenance window.
6. Patch deploymentAdministrators apply the fix across the estate.Deploy in stages; back up first; verify after.
7. Verification / closureThe flaw is confirmed fixed and recorded.Re-scan, update the CMDB, close the ticket.
Exam tip. Know the ordering. The exam likes to ask what happens at each stage, and especially the difference between a zero-day (exploit before stage 5) and an n-day (exploit after stage 5 but before stage 6 completes). The patch gap is stages 5–6.
Part 3

The Cisco IOS software lifecycle

Cisco IOS does not have one version; it has hundreds, organised in a hierarchy you must be able to read. A release train (or just “train”) is a sequence of releases built on the same code base and intended for a particular role. Within a train, individual releases add maintenance fixes and new rebuilds.

Read an IOS version · select a part of 15.2(4)M2

Configuration register · pick a value

PartMeaning
15Major version.
2Minor version (the feature set of the train).
(4)Maintenance rebuild number within the train.
MThe train letter. M = mainline (extended maintenance); T = technology; other letters mark special-purpose trains.
2Rebuild of that maintenance release. Earlier = earlier in the train.

Cisco publishes a latest suggested release (also called the recommended or “validated” release) for every platform and train. Cisco’s guidance is to run the latest validated release for your platform rather than the absolute newest image: it has the most fixed bugs and enough field exposure to be trusted. The Cisco IOS Software Checker maps your running version to known advisories and tells you the suggested fixed release.

Lifecycle milestoneWhat it means
End of Sale (EoS)The product or release can no longer be ordered. Support continues for a while.
End of Software MaintenanceNo new bug fixes or security patches are produced for that release.
Last Date of Support (LDoS)All support ends; the device is fully EOL.
Warning. Running an image past its End of Software Maintenance means a future CVE in that code will never be fixed for you. Planned replacement, not emergency replacement, is the correct response to an EOL announcement.
Part 4

Patch management: the six-step process

Patching is a process, not an event. The standard lifecycle has six steps.

  1. Identify — know what you have and what is vulnerable. Keep an up-to-date asset inventory, scan for software versions, read Cisco PSIRT advisories and CVE feeds, and match them against your estate.
  2. Test — validate the fix in a lab, development, or pilot ring before it touches production. Confirm the device still forwards traffic, routing protocols stay up, and you have a tested rollback.
  3. Schedule — plan a maintenance window. Prioritise by risk: a high CVSS score on an Internet-facing device goes first; an internal, unreachable flaw can wait.
  4. Deploy — apply in stages (a pilot group, then the rest), and for network devices always take a configuration backup immediately before the change.
  5. Verify — confirm the new version is running, services are healthy, and the vulnerability has actually gone. Monitor logs for regressions.
  6. Document — record what was patched, when, by whom, and why; update the configuration management database (CMDB) and the ticket.
Exam tip. The order to memorise is Identify, Test, Schedule, Deploy, Verify, Document. Note that test comes before schedule, and verify comes after deploy — a question that offers “identify, deploy, test…” is wrong by construction.
Part 5

Backing up and upgrading Cisco IOS

Before changing anything, look at what the device already has. The two commands you start with are show version (which image is running and the config register) and show flash: (what images and files are stored).

The upgrade sequence · step through it
1 / 6

      

R1# show version
Cisco IOS Software, C1900 Software (C1900-UNIVERSALK9-M), Version 15.4(3)M2
System image file is "flash0:c1900-universalk9-mz.SPA.154-3.M2.bin"
Configuration register is 0x2102

R1# show flash:
System CompactFlash directory:
File  Length   Name/status
  1   49703812  c1900-universalk9-mz.SPA.154-3.M2.bin
[49703976 bytes used, 62248984 available, 111952960 total]
Exam tip. The pattern to memorise: show flash: to see what is in flash, copy tftp: flash: to pull an image into the device, and copy running-config startup-config to make the configuration (and the new boot system statement) permanent. A configuration that is not copied to startup-config is lost on reload.
Watch your free space. Flash is finite. If show flash: shows too little free space for the new image, delete the old one after you have a working copy elsewhere: delete flash:<old-image>. On some flash types you must run squeeze flash: to reclaim the space. Never delete the image you are currently booted from before the replacement is verified.
Part 6

Image storage, boot, disaster recovery and licensing

Understanding where IOS lives turns an upgrade from a mystery into a checklist. These locations appear throughout IOS commands and exam questions.

LocationName in IOSHolds
ROMrommon / ROM monitorA tiny bootstrap program (and, on some devices, a mini IOS). Survives a wipe; used for recovery.
Flashflash:The full IOS image file (.bin) and often the startup configuration copy. Non-volatile.
RAMrunning-configThe active configuration and the running image. Volatile — lost on reload.
NVRAMstartup-configThe saved configuration. Small and non-volatile.
Externaltftp:, usbflash0:Backups and images held off-device.
Why 0x2142 matters. Setting the register to 0x2142, reloading, and then copying the startup-config back lets an administrator recover from a lost privileged password. An attacker with physical access could do the same — which is why physical security of the device is itself a hardening control (Lesson 4).

Device backup and disaster recovery

Disaster-recovery principle. The two halves of disaster recovery are configuration (restore the startup-config) and software (restore the IOS image). Keep both offsite, keep the image that matches your validated release, and document the exact steps so recovery does not depend on one engineer’s memory.

Licensing: perpetual, subscription, Smart Licensing

Cisco’s licensing has moved from a one-off purchase to a managed, reported model, and the distinction affects whether you can even legally download the patched image. A perpetual licence (the classic Cisco IOS feature-set model, for example universalk9) is bought once and owned forever; it is tied to the image you run and never expires. A subscription licence (for example Cisco DNA Essentials or Advantage) grants the right to use software for a fixed term — typically one, three, or five years — and must be renewed; when it lapses you lose access to the features and to updates. Smart Licensing (today largely Smart Licensing Using Policy) is the mechanism that manages both kinds: the device registers with the Cisco Smart Software Manager (CSSM), reports its usage to a virtual account, and the network stays compliant with the entitlements you own or lease. The practical consequence for this lesson is simple: you can only obtain and run a patched release if your entitlement and support contract are current, so licensing is a security-maintenance dependency, not just an accounting detail. Inspect the state of a device with show license, show license summary, and show license status.

Part 7 · Your turn

Review

Check what you have learned with these review questions.

Review · 6 questions

Sources

Sources & references

  1. Cisco. CCNA (200-301) Exam Topics, Domain 5.0 (Security Fundamentals) and domain 4 (IP services) for device management. cisco.com.
  2. Cisco. Cisco IOS Configuration Fundamentals Command Reference (copy, show flash:, boot system, and the configuration register). cisco.com.
  3. Cisco. Cisco Security Advisories and the Cisco IOS Software Checker (PSIRT). sec.cloudapps.cisco.com.
  4. Cisco. Cisco Product & Software End-of-Life and End-of-Support Policies. cisco.com.
  5. Cisco. Smart Licensing Using Policy configuration guide. cisco.com.
  6. NIST. SP 800-40 Rev. 4: Guide to Enterprise Patch Management Planning (2022). csrc.nist.gov.
  7. FIRST. Common Vulnerability Scoring System (CVSS) Specification. first.org/cvss.

Target exams: CCNA 200-301 v1.1 · CyberOps Associate 200-201. Last content verification: 2026-10-11.