Free Network Engineer Practice Assessment (2026)

Cohesyve · Practice for candidates

Going for a Network Engineer role? Find out how you'd actually score.

Run a Network Engineer simulation built the same way employers build theirs, and get a full report showing exactly where you lost marks — before it counts.

free assessments

no card required

Full

scored report every run

Every

question type included

3 free assessments · No card required · Pro from $16/mo

Overview

Network engineering assessments are dominated by troubleshooting. You are given a topology, a symptom and some output — a routing table, a packet capture, an interface counter — and asked to work back to the cause. Design questions appear too, but the fault-finding sections carry most of the weight because that is what the job is on most days. This page covers the formats and how they are scored.

Why employers assess this role

Network faults are high-impact and time-sensitive, and the difference between an engineer who works methodically down the stack and one who changes things hopefully is measured in outage minutes. Employers assess practically because CV keywords for protocols and vendors are cheap, while structured diagnosis under a clock is visible immediately.

What gets tested

Structured fault diagnosis across the OSI layersIP addressing, subnetting and route summarisationRouting protocol behaviour and path selectionSwitching, VLANs and spanning treePacket capture reading and interpretationFirewall, NAT and access control rule logicNetwork design for redundancy and segmentationChange planning and rollback discipline

The format

Duration

45–90 minutes

Question types

Troubleshooting scenario with device output
Subnetting and addressing calculations
Topology design with redundancy requirements
Packet capture or log interpretation

Levels

Entry · Mid · Senior

What you'll be asked to do

Diagnose a connectivity fault

The largest section in most assessments. A symptom is described, output is provided, and you are scored on the order in which you test things as much as on finding the cause.

•Two hosts on the same subnet cannot reach each other; determine why from switch output
•A site can reach the internet but not the data centre over a site-to-site tunnel
•Traffic to one destination fails only for large packets

Work with addressing and routing

Subnetting is still tested directly and routing questions ask why a particular path was chosen rather than simply which protocol is running.

•Divide an allocated range across six sites with stated host counts and room to grow
•Explain why a route learned from one protocol is preferred over another for the same prefix
•Identify an overlapping subnet that will break routing after a merger

Interpret a capture

A packet capture or interface statistics with a described symptom. Assessors look for whether you read the evidence or pattern-match to a familiar answer.

•Explain a repeated three-way handshake followed by an immediate reset
•Diagnose intermittent slowness from retransmission and window size behaviour
•Identify the cause of increasing interface errors and discards

Design for resilience and segmentation

A smaller but distinguishing section. You design a topology to meet availability and isolation requirements and explain what happens when a component fails.

•Design a branch site that survives the loss of a single link or switch
•Segment a network so that operational technology cannot reach corporate systems
•Plan the failover behaviour for a dual-uplink design and state the recovery time

Cohesyve for candidates

Practise a Network Engineer assessment before the real one

Run the same AI job simulations companies use to evaluate applicants. You get a scored report showing where you're strong and where you're not, plus what to work on.

Sample tasks — and what strong looks like

Users at one site report that some websites load and others time out. All sites use the same internet gateway. Diagnose it.

Mid

What strong looks like: Recognises the selective, size-dependent pattern and tests for a path MTU or fragmentation problem rather than assuming DNS, confirms with a capture or a sized ping, and identifies where the maximum segment size is being broken. Weak answers restart equipment and check DNS repeatedly because that is the familiar script.

Given a routing table and a topology diagram, explain why traffic to a particular prefix is taking an unexpected path.

Mid

What strong looks like: Works through the actual selection order — longest prefix match first, then administrative distance, then metric — and points at the specific entry that wins, rather than describing routing protocols in general. Weak answers explain how the protocol works without ever identifying the winning route.

Divide a /22 across five sites with host counts of 400, 200, 100, 50 and 20, leaving room for growth and keeping summarisation possible.

Entry

What strong looks like: Allocates largest first, sizes each subnet to the next power of two above the requirement, keeps allocations contiguous so they can be summarised, and shows the resulting ranges without arithmetic errors. Weak answers assign equal-sized blocks and run out of space or destroy summarisation.

A core switch is being replaced during a maintenance window. Write the change and rollback plan.

Senior

What strong looks like: Includes a pre-change baseline to compare against, a specific verification step for each stage, a defined point of no return, a rollback that has itself been thought through, and a decision on who calls it. Weak plans list configuration steps with no verification and no way back.

How to prepare

#1
Practise subnetting by hand until you can do it quickly without a calculator, because these questions still appear and slow arithmetic quietly eats the time you need for the troubleshooting sections.
#2
Build a small lab or use a simulator and deliberately break it, then diagnose your own faults, since the habit of working from symptom to evidence is the thing being scored.
#3
Rehearse a structured diagnostic order out loud — physical, then addressing, then routing, then policy — so that under time pressure you do not skip straight to a guess.
#4
Get comfortable reading a packet capture, particularly handshakes, resets, retransmissions and fragmentation, because a capture appears in most mid-level assessments.
#5
Write change plans with explicit verification and rollback steps for changes you have actually made, as this is the section where senior candidates are separated.
#6
Do a complete timed run rather than isolated exercises, since the common failure is spending too long on one fault and leaving the design section blank.

Common mistakes

Jumping to a familiar cause such as DNS without testing the layers below it.
Changing configuration before establishing what the current state actually is.
Explaining how a protocol works in general instead of answering what is happening in the output provided.
Making arithmetic errors in subnetting under time pressure and not sanity-checking the ranges.
Proposing a design with redundant hardware but no statement of failover behaviour or recovery time.
Writing a change plan with no rollback and no verification step.

How it's scored

Criterion	What strong looks like
Diagnostic method	Faults are narrowed systematically with each test chosen to eliminate a specific possibility, and the candidate can say what the result of a test would have meant either way.
Protocol and addressing accuracy	Route selection, subnet boundaries and switching behaviour are described correctly and applied to the specific output given rather than in the abstract.
Evidence reading	Conclusions are anchored to particular lines in the capture, table or counters supplied, not to what the symptom usually means.
Design and resilience	Redundancy is paired with a stated failover mechanism and recovery time, and segmentation is enforced where it cannot be bypassed.
Change discipline	Baselines, verification points, rollback and communication are all present, and the plan is written so another engineer could execute it.

Frequently Asked Questions

Will the assessment be vendor-specific?

It depends on the employer, and it is normally stated in advance. Many assessments deliberately use generic output so that protocol understanding rather than command syntax decides the score, and fundamentals transfer between vendors far better than memorised commands.

Is subnetting still tested?

Yes, regularly, and usually under time pressure without a calculator. It is one of the few areas where pure practice reliably improves a score, so it is worth drilling even if you use tooling day to day.

How much cloud networking comes up?

Increasingly often — virtual networks, peering, hybrid connectivity and cloud security groups appear in many assessments now, particularly for roles at organisations running a hybrid estate. The underlying routing and segmentation reasoning is the same.

What separates a mid-level score from a senior one?

Senior candidates are marked on change planning, blast radius and communication as much as on diagnosis. Finding the fault is expected; explaining the risk of the fix and how to back it out is what moves the score up.

How should I practise before sitting one for real?

Practise under marking, not just in a lab. Cohesyve offers three free scored assessments with a full report, which will show you whether your diagnostic method reads as systematic to someone else.

Practise another role

Database Administrator →Solutions Architect →IT Support Specialist →Product Manager →

Cohesyve · Practice for candidates

Practise a Network Engineer assessment now — free.

Three free scored assessments, a full report on every run, and a learning pathway built from what you got wrong. No card required.