Reviewed 10 October 2026 against the linked published scope. V15 domains: mobile devices 13%, networking 23%, hardware 25%, virtualization/cloud 11%, troubleshooting 28%.
Memory hook: Fit, connect, power, configure, then test the original symptom.
Read the essentials, cover the answers and explain the decision aloud. Open the topic summaries below whenever a distinction is unclear.
Mobile and networking essentials
- Before servicing a mobile device, protect data, disconnect power and follow battery/ESD precautions. Swollen or damaged lithium batteries require the approved handling process; do not puncture or force them back into place.
- Match RAM form factor, supported generation, storage interface and physical dimensions. A connector that fits is not proof that a cable supports the required display, speed or charging standard. USB-C is a connector; USB, Thunderbolt and Power Delivery describe capabilities.
- Docks combine power, network, display and peripherals subject to device support. Screen brightness, background apps, wireless radios and failing batteries affect runtime. MDM controls enterprise enrollment, configuration and remote actions; BYOD ownership changes policy boundaries.
- An AP provides wireless access, a switch joins local Ethernet devices, a router connects IP networks and a firewall applies traffic policy. DHCP provides address configuration; DNS resolves names; NAT translates addresses.
- Memorize commonly tested ports: SSH/SFTP 22, DNS 53, HTTP 80, HTTPS 443, SMB 445, RDP 3389; SMTP 25, POP3 110, IMAP 143 and their protected options. Know when a protocol sends credentials without suitable protection.
- Use a client IP, subnet, gateway and DNS check to separate link, addressing, routing and name problems. 169.254.x.x often signals failed DHCP acquisition. WPA2/WPA3, strong administration credentials and firmware maintenance secure a small network.
Hardware, storage and cloud
- CPU socket, chipset, firmware, power delivery and cooling must support the processor. DIMM versus SODIMM, memory generation, capacity and ECC support must match the platform. More GHz alone does not establish overall performance.
- SATA is a storage interface; NVMe commonly uses PCIe. M.2 describes a form factor, so verify keying, length and supported protocol. HDDs have moving parts; SSDs still fail and need backups.
- RAID 0 stripes without redundancy; RAID 1 mirrors; RAID 5 uses single distributed parity; RAID 6 uses double parity; RAID 10 stripes mirrored sets. RAID availability is not protection from deletion, malware or every correlated failure.
- Power supplies must match wattage, connectors, input and form factor. A UPS supplies temporary power; a surge protector does not provide runtime. Never open a power supply to service its internal high-voltage components.
- Virtual machines share physical resources through a hypervisor; allocate CPU, RAM, disk and networking deliberately. Containers share a kernel, unlike full guest VMs. Snapshots assist rollback but are not independent backups.
- IaaS supplies infrastructure, PaaS an application platform and SaaS an application. Public/private/hybrid cloud describe deployment arrangements. Elasticity changes capacity; high availability reduces disruption; synchronization is not archival backup.
Printers and fault isolation
- Laser sequence: processing, charging, exposing, developing, transferring, fusing, cleaning. Toner that rubs off points toward fusing; repeating marks can indicate a damaged rotating component. Follow the model's diagnostics before ordering parts.
- Inkjet uses liquid ink and may need nozzle cleaning/alignment; thermal printers need suitable media and heating; impact printers use ribbon and mechanical strike. Duplexing, trays, drivers, permissions and spoolers are separate fault sources.
- No power: verify supply, outlet, cable, switch and connectors. Power but no POST: inspect firmware indicators, RAM seating and hardware compatibility. Overheating: inspect airflow, fans, heatsink contact and dust.
- For storage faults, protect data first, inspect cabling/health and distinguish missing boot media from filesystem or OS problems. For display faults, check input selection, cables, brightness, adapters and a known-good display.
- Work in order: identify, establish a theory, test, plan, implement, verify and document. Change one relevant factor at a time; compare before/after behavior rather than declaring success from a component light.
Final active recall
1. Does an M.2 SSD necessarily use NVMe?
No. M.2 devices may use different protocols; verify the socket and drive specifications.
2. Does RAID replace a backup?
No. It can improve availability but does not preserve independent recovery history.
3. Laptop cannot charge from a cable that transfers files. Why?
Cable, charger and device may not negotiate or support the required power delivery.
4. Laser text rubs off the page. Which stage is implicated?
Fusing, which bonds toner to paper; confirm with the printer diagnostic process.
5. A client has an APIPA address. What should you investigate?
The link/VLAN and DHCP path, including server availability and relay configuration.
Sources and further practice
- Official exam scope
- Objective-to-topic coverage map. Each full topic links to its supporting primary technical documentation.
Every topic at a glance
Open any topic to revisit its essential facts, decisions and exam traps. Use the full topic for active recall and supporting references.
01 · Mobile Devices and Laptop Support
Memory hook: Port shape is not capability; synchronization is not backup.
Must remember
Laptop repair starts with the service manual, compatible parts, power isolation and ESD precautions. Batteries, keyboards, storage, RAM, wireless antennas and displays may be replaceable or integrated depending on the model. A swollen lithium battery requires safe handling and escalation; do not puncture, compress or keep charging it.
Distinguish LCD backlight/display faults from GPU or connection faults by comparing internal and external output, brightness, artifacts and cable behavior. Touch digitizers, webcams, microphones and biometric sensors are separate components even when integrated into one assembly. Antenna placement/cables can affect Wi-Fi after screen service.
USB-C describes a connector shape, not guaranteed charging power, video output or transfer speed. Check the host, cable, dock and peripheral capabilities, including USB/Thunderbolt/DisplayPort alternate modes where supported. A cable may charge successfully yet lack the data/video capability required. Docks can add displays, networking and peripherals while depending on firmware, power delivery and OS drivers.
Wi-Fi provides local wireless networking; Bluetooth pairs nearby peripherals; NFC supports very short-range interactions; cellular access uses the mobile network and an appropriate SIM/eSIM/account. Airplane mode, roaming, hotspot settings, metered data and carrier coverage affect connectivity differently.
Synchronization keeps selected data consistent across services/devices; it can also propagate deletion. Configure account, mail/calendar/contact settings, permissions and conflict handling. MDM may enforce encryption, lock, remote wipe and application policy; BYOD requires attention to the boundary between personal and organizational data.
Troubleshoot battery drain, heat, weak signal and unavailable peripherals using recent changes, settings, logs and known-good compatible accessories before replacing the whole device.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Dock charges but no monitor output | Check video support across host, port, cable and dock. |
| Battery case bulging | Stop use/charging and follow safe manufacturer handling. |
| Deleted contacts disappear everywhere | Synchronization propagated the change; seek a separate recovery history. |
Traps
- USB-C does not imply every cable supports every feature.
- A synchronized folder is not necessarily an independent backup.
02 · IPv4, IPv6 and Subnetting
Memory hook: Prefix is the boundary; longest match wins.
Must remember
IPv4 has 32 bits. CIDR /n reserves n network bits, leaving 32-n host bits. A conventional IPv4 subnet has 2^(32-n) addresses and usually two reserved endpoints. /31 point-to-point and /32 host routes are exceptions to the usual minus-two rule.
Worked example: 192.168.10.77/26 has blocks of 64: network .64, broadcast .127, ordinary hosts .65–.126, 62 usable. Four /26 networks fit within a /24. For VLSM, allocate the largest required blocks first and check overlap.
| Range | Meaning |
|---|---|
| 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16 | RFC1918 private addresses; not public Internet destinations. |
| 169.254.0.0/16 | IPv4 link-local/APIPA; often a DHCP failure clue. |
| 127.0.0.0/8 | IPv4 loopback. |
| 224.0.0.0/4 | IPv4 multicast. |
| ::1 / :: | IPv6 loopback / unspecified address. |
| fe80::/10 | IPv6 link-local. |
| fc00::/7 | IPv6 unique-local range. |
| ff00::/8 | IPv6 multicast. |
IPv6 uses 128 bits in hexadecimal groups. Remove leading zeros and compress one consecutive run of zero groups with ::; using it twice would make the address ambiguous. Many LAN subnets use /64, but do not assume every IPv6 prefix is /64.
ARP discovers IPv4 next-hop MAC addresses on a local link. IPv6 Neighbor Discovery uses ICMPv6; blocking all ICMPv6 can break normal operation. Router advertisements supply on-link/default-router information and can support SLAAC. DHCPv6 does not replace every router-advertisement function.
Historical class A/B/C boundaries are vocabulary; current route decisions use actual prefixes. Two hosts with mismatched masks can disagree about whether traffic is local. Verify address, prefix, gateway and duplicate-address evidence together.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Need 50 ordinary IPv4 hosts | A /26 provides 62 usable addresses. |
| Host has 169.254 address unexpectedly | Investigate DHCP/link configuration. |
| IPv6 address works locally but not remotely | Check router advertisements/default route and policy. |
Traps
- 172.16/12 covers 172.16 through 172.31, not all 172 addresses.
- A default gateway must be reachable through the host’s local routing configuration.
03 · Ports, DNS, DHCP and Time
Memory hook: Resolve the name, obtain the lease, trust the clock.
Must remember
| Service | Conventional port(s) |
|---|---|
| FTP / SSH-SFTP / Telnet | TCP 21 control (FTP data varies) / 22 / 23. |
| SMTP / submission | TCP 25 / 587; implicit TLS submission commonly 465. |
| DNS | UDP and TCP 53. |
| DHCPv4 / DHCPv6 | UDP 67 server, 68 client / 547 server, 546 client. |
| HTTP / HTTPS | 80 / 443; HTTP/3 uses QUIC over UDP 443. |
| POP3 / IMAP | TCP 110 / 143; implicit TLS 995 / 993. |
| SNMP queries / traps | Usually UDP 161 / 162. |
| LDAP / LDAPS | 389 / 636. |
| SMB / RDP | TCP 445 / TCP and UDP 3389. |
| NTP / SIP | UDP 123 / commonly 5060, TLS 5061. |
Ports are conventions, not proof of application identity. Secure FTP (SFTP over SSH) differs from FTP with TLS (FTPS).
DNS A/AAAA map names to IPv4/IPv6; CNAME aliases another name; MX specifies mail exchangers; NS delegates authoritative service; PTR supports reverse lookup; TXT carries text used by policies and verification. A resolver asks authoritative servers and caches answers according to TTL. DNSSEC authenticates signed DNS data; it does not encrypt ordinary DNS queries. Encrypted DNS protects the resolver connection through its selected protocol.
DHCPv4 commonly follows Discover, Offer, Request, Acknowledge. Scopes define pools, options, exclusions and reservations. A relay forwards requests across routed boundaries. Exhausted pools, rogue servers, wrong options and blocked relay paths produce different symptoms. Reservations provide predictable leases, not static manual configuration on the host.
NTP synchronizes time broadly; PTP supports more precise timing in suitable networks; NTS adds authenticated security to NTP. Time matters for logs, certificates and ticket-based authentication.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Name resolves to wrong address | Check authoritative record, cached answer, TTL and split-DNS view. |
| Clients across router cannot obtain leases | Inspect DHCP relay and server scope. |
| Need trustworthy signed DNS answers | DNSSEC; use separate transport protection if confidentiality is required. |
Traps
- DNS can use TCP, not only UDP.
- A DHCP reservation does not prevent every rogue device from using an address.
04 · Troubleshooting Networks
Memory hook: Define the symptom; test one layer; verify the whole path.
Must remember
Start by identifying scope, symptoms, recent changes and what still works. Establish a plausible theory, test it, plan a low-impact fix, implement under the change process, verify full functionality and document cause/results. Escalate when the evidence or authority requires it.
| Symptom | Useful next check |
|---|---|
| No link | Cable/optic, administrative state, speed and power. |
| Increasing CRC errors | Physical quality, interference, optic/cable mismatch. |
| Link up, no usable address | VLAN and DHCP/relay/pool state. |
| Local works, remote fails | Prefix, gateway, route and ACL. |
| IP works, name fails | Resolver settings, DNS records and cache. |
| Small traffic works, large transfers stall | MTU/path-MTU and filtered ICMP. |
| Wireless intermittent | Channel contention, roaming, SNR and retries. |
ipconfig/ip inspect host configuration; ping checks selected reachability; traceroute/tracert expose responding hops; nslookup/dig query DNS; arp/ip neigh show neighbor bindings; netstat/ss show sockets; tcpdump or Wireshark inspect packets. Command availability differs by OS. A blocked ICMP response or a router deprioritizing probes does not necessarily mean the application path is broken.
Use cable testers for continuity/wiremap, toner/probe tools for cable identification, optical meters/OTDR for suitable fiber diagnosis and Wi-Fi analyzers for channel/radio evidence. Choose the instrument for the suspected fault; a speed test does not identify every wiring defect.
Inspect both directions. A request can leave successfully while the reply lacks a route, session state or policy permission. Duplicate addresses, wrong masks, exhausted DHCP pools, native-VLAN mismatches and stale DNS can appear intermittent.
Finish by testing the actual user service from the intended source, not just the nearest gateway. Record the fixed cause and update diagrams/baselines when the intended design changed.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Only one user fails | Compare that client/port/configuration with a working peer. |
| Everything fails after a change | Correlate the changed dependency and use the approved rollback if warranted. |
| Ping succeeds but web request fails | Test DNS, transport port, TLS and application behavior. |
Traps
- Traceroute timeouts can reflect filtering rather than a failed forwarding hop.
- Changing several variables at once destroys diagnostic clarity.
05 · Motherboards, CPUs, Memory and Firmware
Memory hook: Socket, generation, power and firmware must agree.
Must remember
Motherboard form factor affects chassis fit, mounting and expansion. CPU compatibility requires the correct socket, chipset support, power/thermal design and firmware version; matching only the socket is insufficient. Integrated graphics availability depends on the CPU/platform, not simply the presence of a display connector.
RAM modules differ by generation, form factor, voltage and platform support. DIMMs commonly serve desktops; SODIMMs many laptops; some systems use soldered memory. DDR generations are not interchangeable. ECC detects/corrects supported memory errors and requires compatible platform support. Populate channels/slots according to the manual rather than guessing from color alone.
UEFI firmware initializes hardware and boots an OS. Secure Boot validates an approved boot chain; TPM-backed capabilities support key protection and measured trust. Firmware passwords, boot order, virtualization extensions and storage mode can affect installation and recovery. Record settings before changes; changing storage mode can prevent an existing OS from booting.
Cooling depends on correct heatsink contact, suitable thermal interface material, fan/pump function and unobstructed airflow. Thermal throttling can cause slow performance before shutdown. Avoid mixing incompatible modular PSU cables even if connectors appear to fit: pinouts can differ.
Select a power supply for total demand, connector types, efficiency and quality with appropriate headroom. Desktop rails/components use specific voltages; do not open power supplies or CRT equipment because retained high voltage is hazardous. Use proper ESD grounding for component work, but electrical safety and power isolation come first.
Read POST codes/beeps and diagnostic LEDs using the model's documentation. A failure before OS load points toward firmware/hardware, while an OS crash can have either hardware or software causes.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Upgrade CPU | Verify socket, chipset/firmware support and power/cooling. |
| Random errors after RAM replacement | Check compatibility, seating, channel population and diagnostics. |
| Slow only under sustained load | Inspect temperatures, throttling and power limits. |
Traps
- An identical-looking connector does not guarantee compatible PSU wiring.
- More RAM cannot fix every CPU, disk or thermal bottleneck.
06 · Storage, Displays and Peripheral Connections
Memory hook: Interface, protocol and form factor are different questions.
Must remember
HDDs use moving magnetic media; SSDs use flash. SATA is a storage interface/protocol family; NVMe uses a PCIe-based storage protocol. M.2 is a form factor that can carry different interfaces, so an M.2 drive and slot must agree on supported protocol, length and keying.
RAID 0 stripes for capacity/performance without redundancy; RAID 1 mirrors; RAID 5 uses distributed single parity; RAID 6 tolerates two drive failures; RAID 10 combines mirroring and striping with failure tolerance depending on which members fail. RAID is not a backup: deletion, corruption or a controller/site failure can affect the array. Rebuild time and degraded performance matter.
SATA data/power, PCIe slots, USB, Thunderbolt, HDMI, DisplayPort, Ethernet and audio connections carry different capabilities. An adapter may be passive or require active conversion; direction and resolution/refresh support matter. Display troubleshooting includes input source, cable, resolution, refresh rate, scaling, GPU output and monitor controls.
Bluetooth/USB peripherals can fail through pairing, drivers, power management, permissions or interference. Compare behavior on another supported port/device using a known-good cable. A working charger does not validate every data conductor.
Storage symptoms include SMART warnings, unusual HDD noises, read errors, low free space and high latency. Preserve important data before stress testing a potentially failing disk. A filesystem error and physical failure can coexist. Use appropriate diagnostics and verify restores when replacing drives.
Plan storage by capacity, random/sequential workload, endurance, latency, redundancy and recovery requirements rather than only advertised peak transfer rate.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| M.2 drive not detected | Check SATA versus NVMe support, slot/keying and firmware. |
| Need redundant disks | Choose an appropriate RAID design plus independent backups. |
| Display works at low but not high refresh | Check cable/adapter/bandwidth and supported mode. |
Traps
- M.2 does not automatically mean NVMe.
- RAID cannot undo accidental deletion.
07 · Printers, Scanners and Print Troubleshooting
Memory hook: Identify the print technology before replacing a part.
Must remember
Laser printing uses a charged imaging system and toner. Remember the sequence conceptually: processing, charging, exposing, developing, transferring, fusing and cleaning. A fuser bonds toner using heat/pressure; a drum/imaging surface is different from the fuser. Repeated marks, loose toner and faded output point to different stages and consumables.
Inkjet printers spray liquid ink; clogged nozzles, dried cartridges, alignment and paper/media compatibility affect output. Thermal printers use heat-sensitive media or transfer ribbons depending on design; incorrect paper or orientation can produce blank receipts. Impact printers use pins/ribbons and can handle multipart forms, with mechanical/ribbon wear patterns.
Check paper size/type, tray selection, duplex path, rollers, jams, consumables, duty cycle and maintenance kit guidance. Use approved cleaning procedures; hot fusers and moving parts require safe handling. A scanner's glass/feeder can cause streaks unrelated to the printer engine.
The print path includes application, driver, spooler/queue, network/USB connection and device. Print a device self-test to separate engine faults from host/driver/network issues. A stuck queue, incorrect default printer, wrong driver/language or changed IP can prevent output despite a healthy engine.
Network printing needs stable addressing/discovery and appropriate permissions. Secure print release protects unattended documents; duplex and sensible defaults reduce waste. Shared printers and direct-IP printing have different host dependencies.
Verify the user's original document and finishing requirements after a fix. A basic test page may not exercise the failed tray, duplex unit, color path or scanner.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Toner wipes off the page | Inspect fusing/media settings and relevant hardware. |
| Printer self-test works but PC job fails | Investigate driver, queue, permissions and connection. |
| Vertical line only when using document feeder | Inspect/clean the relevant scanner strip and feeder path. |
Traps
- A drum and a fuser are not interchangeable parts.
- Repeatedly resending a stuck job can create a large duplicate queue.
08 · Virtual Machines and Cloud Basics
Memory hook: Virtual hardware still consumes real resources.
Must remember
A hypervisor presents virtual hardware to guest operating systems. Type 1 runs directly on the host platform; Type 2 operates through a host OS. VMs isolate guest kernels; containers usually share a host kernel. Neither removes the need to patch, authorize access or back up data.
Plan CPU virtualization support, memory, storage and network capacity for concurrent guests. Overcommitting memory or disk can harm both host and guests. Virtual switches/network modes determine reachability: NAT, bridged/external and host-only/internal modes solve different needs. Do not expose a practice VM to a production network unintentionally.
Snapshots/checkpoints capture a recoverable state under the platform's model; they can grow and are not a substitute for independent backups. Clones create separate instances whose identities, licenses and credentials need review. VDI delivers centrally hosted desktops; local virtualization runs guests on the endpoint.
Cloud IaaS provides infrastructure, PaaS a managed application platform and SaaS a managed application. Public/private/hybrid/community describe deployment arrangements. Elasticity responds to changing demand; scalability is the ability to grow; high availability reduces outage impact.
Shared responsibility changes operational duties, but customer data, identities and configurations still need protection. Pay-as-you-go shifts cost patterns without guaranteeing savings. Network dependence, data location, subscription licensing and backup/recovery remain part of support.
Common endpoint uses include testing, legacy applications, isolated development and training. Select a supported, licensed environment and verify both guest functionality and host health.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| Test another OS without replacing the main OS | A suitably resourced virtual machine. |
| Centrally delivered corporate desktops | VDI or an appropriate desktop-as-a-service solution. |
| Use a completed web application | SaaS rather than managing its underlying VM stack. |
Traps
- A snapshot stored with the VM can be lost with the same storage failure.
- Virtualization isolation is not permission to run untrusted code without controls.
09 · Hardware Diagnosis and Safe Verification
Memory hook: Protect data, isolate the cause, prove the fix.
Must remember
Identify the problem and scope, ask about recent changes, establish a theory, test it, plan the fix, implement under authorization, verify and document. Start with simple evidence: power, connections, input selection and error messages. Do not replace expensive parts before checking a loose cable.
No power differs from no POST, no boot and no display. Verify outlet/adapter/power indicators, then POST diagnostics, firmware device detection and boot configuration. A black screen may still come from a running system with the wrong display path. Intermittent shutdown under load suggests heat or power but requires measurements.
Use known-good compatible components to isolate faults. Memory diagnostics, storage health data and vendor self-tests support a theory; one passed test cannot prove absence of every intermittent fault. Multimeters require appropriate training/range and electrical-safety precautions; never probe hazardous equipment casually.
Network failures can originate at the physical link, VLAN, address, gateway, DNS or application. A loopback plug tests a limited path, while a cable tester checks wiring/continuity; neither proves end-to-end service success. Wi-Fi analyzers expose radio/channel evidence.
Preserve data and recovery keys before storage, firmware or OS changes. Document serials, replaced parts, settings and results. Dispose of batteries/electronics through approved processes. Use ESD protection and lifting practices, and keep liquids/dust away from sensitive equipment.
Finish by reproducing the user's original workflow and checking side effects. Explain the change in plain language, obtain required acceptance and update the ticket so the next technician can understand the evidence.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| System powers but does not POST | Use board/vendor diagnostics and minimum compatible hardware isolation. |
| User files at risk | Prioritize approved backup/recovery before destructive repair. |
| Intermittent fault | Correlate load, temperature, power and logs rather than guessing. |
Traps
- No display is not identical to no power.
- A successful reboot is not complete verification of every affected function.