CCNA deep review — September 29, 2026

The guide maps all 53 current numbered objectives and their supporting bullets from the actual four-page v1.1 PDF, distributed 13/9/5/9/10/7. The actual three-page v2.0 PDF has 29 numbered objectives, distributed 7/5/4/7/6; its future map is retained separately. Both PDFs were read in full and binary/page receipts preserved. These receipts do not establish a binary change from a previously observed file.

The canonical exam-page objective and lifecycle digests are unchanged. Its HTML does not itself contain the full PDF scope or transition notice, so an unchanged monitor is not sufficient evidence for the full review. No objective snapshot was rewritten. Ryan Rose’s May 20, 2026 announcement and July 1 transition article establish the last current exam on February 2, 2027 and the new exam from February 3. Expired promotions and other certifications’ changes were not adopted as CCNA requirements.

Scope and technical corrections

The previous transition paragraph incorrectly grouped DHCP with topics leaving the named future scope. DHCP moves to domain 1 as troubleshooting of client, server and relay. SNMP and syslog remain in domain 5. The guide preserves current v1.1 learning and separates future configuration/troubleshooting, OSPFv3, DNS records, secure file transfer, Layer 2 controls and AI/Ansible tasks. The future OSPF adjacency objective explicitly excludes authentication.

Cisco’s route-selection explanation supports a distinction between candidate selection/route installation and later packet forwarding. The revised guide explains why a more-specific installed prefix wins despite a higher AD and why unrelated protocol metrics are not common units. Floating-static behavior depends on route eligibility and withdrawal, not merely an application outage. The local workbook exercises those distinctions without claiming next-hop probes or real convergence.

LACP member loss is now conditioned on minimum-links requirements; reduced capacity is not guaranteed when the whole bundle becomes ineligible. OSPF teaching separates normal broadcast DROTHER 2-WAY relationships from persistent database-exchange faults; MTU mismatch is one possible cause, not a reason to blindly ignore MTU. STP role/state teaching now includes learning and reconciles generic RSTP terminology with platform blocking/disabled labels, without transplanting NX-OS configuration into IOS.

ACL material adds noncontiguous wildcard matching, rule shadowing, independent return-direction policy and ACK/RST matching without a connection-state database. DAI teaching distinguishes DHCP-learned bindings, static-host ARP ACLs and trusted interfaces that bypass checks. API material adds pagination and error handling without suggesting every Cisco API has the Secure Endpoint contract. Worked subnet and modified EUI-64 examples were separately checked. Technical documents were read in selected sections, with precise boundaries in the source notes; this was not a comprehensive review of every linked configuration guide or RFC.

Executed local evidence and remaining labs

The exact public Python workbook passed 48 checks. It builds six synthetic installed prefixes, retains two equal-cost OSPF candidates, changes to a floating static after supplied eligibility changes, restores the original candidates and preserves a more-specific host route. It checks IPv4/IPv6 separation, no-route cases, unusable candidates, AD 255, invalid/tied model inputs and input-order independence. ACL cases test permitted/denied/unmatched traffic, wildcard boundaries, reversed ordering, a denied reply, ACK/RST behavior and a JSON round trip.

The code is an original bounded teaching model. Its route eligibility is provided, not measured. It supports one process and comparable metric class per source and rejects unsupported AD combinations. It implements neither real OSPF nor recursive next-hop resolution, tracking, convergence, policy routing or ECMP hashing. ACLs use selected parsed IPv4 fields and trusted fixtures, not IOS syntax or packet parsing, fragmentation, NAT, interface application, comprehensive validation or session tracking. Some checks are structural/input-rejection assertions. Four separate checks validate subnet/range/host-count and EUI-64 arithmetic.

There was no device, simulator, wireless controller, authenticated API, credential, socket, provider lab or live enforcement exercise. The guide now contains 48 individually answered original prompts, three retained integrated scenarios and eight concrete proposed labs with controlled faults, expected observations and restoration. These labs and independent human review remain pending.

Logistics and learning resources

The canonical exam page still lists 120 minutes, USD300 and English/Japanese. The credential page still lists USD400 and English only. The central pricing table corroborates USD300 but has stale version labels elsewhere. The guide retains the discrepancy and recommends checking the actual scheduling checkout, without inventing a reconciliation. The dedicated recertification page lists Associate renewal through 30 CE credits and requires completion during the active period; no account balance or eligibility was checked.

The Pluralsight path has 22 public course cards totaling 3,461 minutes, or 57 hours 41 minutes, versus a 58-hour header. Most dates are 2020–2021, with two August 2024 updates and a September 11, 2026 IPv4 course. The old 59-hour estimate was replaced; listing dates do not prove complete current or future coverage.

Cisco Press lists a July 25, 2024 second-edition two-volume library by Wendell Odom, David Hucaby and Jason Gooley, companion practice and an update program. Its advertised eight-plus video hours are not overall book study time. NetAcad’s landing page verifies 22-hour Networking Basics and a two-hour Packet Tracer introduction, not a full CCNA curriculum total. CML advertises a five-node free experience and version 2.10 features including beta wireless; no entitlement or lab compatibility was verified.

O’Reilly and Udemy returned 403. YouTube exposed a short title/footer shell. Prior durations and update claims from those routes are explicitly unverified. No paid lessons, book interiors, proprietary questions or subscriber lab instructions were accessed. The resource table is last, uses the required three columns and distinguishes provider durations from editorial practice budgets.

Operational evidence preserves the full current/future mappings, actual PDF receipts, exact workbook/runtime/output, catalog arithmetic, source reading limits, unchanged monitor observations and repository gates. No reserved GitHub guide or notification setting was changed.