H20-923_V1.0 HCSP-Field-Data Center Facility V1.0 Question and Answers

Smart cooling temperature and humidity control is based on where the control “reference point” is measured and regulated.Supply air controluses the unit outlet sensor as the main reference, so the system directly regulates the temperature (and humidity when configured) of the air delivered to the IT area. This provides stable supply conditions and is widely used for predictable airflow organization.Return air controluses the unit inlet/return sensor as the reference, reflecting the heat absorbed from IT loads; it helps the unit respond to real load changes and maintain stable operation when room mixing or load distribution varies.Cold aisle controluses sensors placed in the cold aisle (near rack inlets) as the reference, aligning control with the most critical point for IT reliability—server intake conditions—especially in contained aisle scenarios.Hot aisleis generally not used as a primary temperature-and-humidity control mode because hot aisle conditions are intentionally high and variable (driven by load and containment), and controlling to hot aisle targets can conflict with ensuring stable rack inlet temperature and proper dehumidification behavior. Therefore, A, B, and C are valid control modes.

In Huawei alarm management,masking rulesare used to control how alarms are handled and displayed during specific conditions, such as planned maintenance, commissioning, device replacement, or known temporary abnormalities. When a masking rule is applied, the system can change how an alarm is classified so that O&M staff can distinguish between alarms that require immediate action and alarms that are expected or should be excluded from operational statistics. Two key statuses that masking rules can assign areMaintenanceandInvalid.Maintenanceindicates the alarm is generated during an approved maintenance activity (for example, power-off tests, sensor rewiring, cooling maintenance), so it is treated as expected and does not trigger normal escalation logic.Invalidindicates the alarm is not considered effective for operational follow-up—commonly used when a point is confirmed to be non-applicable, temporarily disabled, or its triggering condition is not meaningful to current operations. By contrast,ValidandStandardrepresent normal alarm handling states rather than outcomes specifically “identified by masking rules.”

For the Huawei UPS5000-H, inverter startup is executed through theUPS’s own control interfaces, which provide the “Running Control / Inv. ON” operation with the required confirmations and permissions. TheUPS LCD (MDU touchscreen)supports inverter startup through menu operations such as selecting inverter start functions and confirming the action after authentication. TheUPS WebUIalso provides the same operational control path, where an operator logs in and issues theInv. ONcommand under the UPS monitoring/running control pages. These two interfaces are part of the UPS monitoring and control design and are explicitly used for commissioning and routine O&M actions.

By contrast,ECC(data center controller) andNetEco(management platform) are mainly used for site-level monitoring, alarm collection, visualization, reporting, and centralized management. While they can display UPS status and alarms (and may integrate with UPS data), inverter startup is defined as aUPS running control operationperformed on the UPS’s LCD or WebUI control plane to ensure proper authorization, safety checks, and command traceability.

This statement is true. In split DX/precision cooling refrigerant systems, the “gas pipe” typically refers to thesuction linereturning refrigerant vapor (and entrained compressor oil) from the indoor evaporator to the outdoor unit/compressor. When theoutdoor unit is higher than the indoor unit, the suction line includes avertical riser. Oil naturally tends to settle and flow downward under gravity, so without proper piping measures, oil may accumulate in low points or fail to be carried upward consistently, leading topoor oil return. Insufficient oil return can cause compressor lubrication issues, abnormal noise, overheating, reduced efficiency, and long-term reliability risks. Installingoil trapsat the bottom of the riser and thenevery 5–6 metersof vertical lift helps collect oil and allows it to be periodically swept upward by refrigerant velocity, maintaining stable oil circulation. This requirement is especially important in data center cooling because continuous operation and high availability demand reliable compressor lubrication and stable refrigerant flow under varying load conditions.

Routine UPS maintenance is preventive by design, so it must cover bothenvironmental conditionsandequipment operating status, not only alarms. The equipment room environment directly affects UPS reliability: high temperature accelerates capacitor aging, reduces battery/rectifier reliability, and may trigger derating; improper humidity increases the risk of condensation or electrostatic discharge; dust and poor housekeeping raise the risk of blocked airflow and overheating. Therefore, theenvironment checkin option A is a standard routine item. Option B is also essential because O&M personnel must confirm the UPS is innormal operating mode, verify no hidden alarms exist in the event log, and ensure the load level is within the recommended range to maintain redundancy and efficiency. Option C is true because fans are typical “wear-out” components; abnormal noise or speed changes often appear before a failure and can be detected early through inspection. Option D is false: absence of a front-panel alarm never replaces onsite environmental inspection and basic preventive checks.

Huawei’s scenario-based data center facility portfolio is organized around clearly defined solution families that map to typical deployment scenarios and site sizes. FusionModule500 and FusionModule2000 are smart modular data center solutions designed for standardized, fast deployment in equipment rooms and enterprise edge/branch scenarios.They integrate key subsystems such as power supply and distribution, cooling, monitoring, and optional fire protection into a modular architecture that reduces footprint and accelerates on-site delivery. Likewise, FusionDC1000 is Huawei’s prefabricated modular data center series, delivered as complete, factory-integrated modules that combine civil engineering elements with electromechanical infrastructure for rapid construction, predictable quality, and scalable expansion across different capacity tiers. In contrast, FusionDC800 is not used as a defined scenario-based solution name within Huawei’s mainstream data center facility solution lineup (where “FusionDC” offerings are represented by the FusionDC1000 series and “FusionModule” offerings by numbered FusionModule solutions). Therefore, FusionDC800 is the option that does not match Huawei’s scenario-based solution naming.

Compressor preheating is a protection mechanism used during first power-on (and after long power-off periods) to improve compressor reliability before allowing refrigeration startup. The preheating function warms the compressor oil and reduces refrigerant migration and oil dilution, which otherwise can lead to liquid refrigerant in the compressor crankcase. If the compressor starts when oil is diluted or when liquid refrigerant is present, it increases the risk of abnormal noise, poor lubrication, high mechanical stress, and even liquid strike, which can damage the compressor and shorten service life. For data center cooling equipment, where continuous reliability is essential, the commissioning logic treats preheating as a required prerequisite for safe compressor operation. Therefore, O&M practice does not recommend bypassing or manually shutting down compressor preheating just to accelerate commissioning. The correct approach is to keep the unit energized and allow the preheating period to complete, then proceed with commissioning steps (fans, cooling system tests, alarms, and control verification) under proper protection conditions.

Free cooling is a mainstream energy-saving concept in data center cooling design, used to reduce chiller compressor runtime by leveraging low outdoor temperatures (air-side) or low cooling-water temperatures (water-side) during suitable seasons. Therefore, describing the technology as “not mature enough and should be used with caution” is the false statement when discussing its advantages. In practice, free cooling has broad industry adoption and is considered a proven approach when applied with correct engineering, such as proper filtration, anti-condensation control, heat exchanger selection, and automatic control strategies. Its advantage is that it can directly utilize the outdoor “cold source” to remove heat, which significantly reduces cooling power consumption compared with full mechanical refrigeration, improving overall energy efficiency and supporting lower PUE targets. It also makes full use of seasonal and day-night temperature differences: the system dynamically switches among free cooling, partial free cooling, and mechanical cooling to balance reliability and efficiency. The key point is maturity: the benefit of free cooling comes from well-established principles and widely implemented architectures, not experimental technology.

As data center power density increases (especially driven by AI and high-performance computing), the industry trend is to adoptliquid coolingin high-density scenarios because liquid has much higher heat transfer capability than air, enabling stable cooling where rack power is very high. This makes statementAtrue. The combination ofCloud + AIpushes operators to reduce PUE and total energy cost through more intelligent control, higher efficiency cooling architectures, and integrated energy management, so statementBis also true. For large data centers, another clear trend is increasingchilled-water supply/return temperatures(higher inlet and outlet temperatures). Warmer water improves chiller efficiency, expands the number of hours that free cooling can be used, and reduces compressor work, soDis true. However,evaporative coolingis not universally the “main solution” for large data centers because it is climate- and water-condition dependent and is typically deployed as one option among multiple architectures rather than the dominant default everywhere. Therefore,Cis the incorrect statement.

In Huawei smart module monitoring architecture, theUIM20A expansion moduleis traditionally used toincrease device portsso cabinets can connect multiple intelligent components and sensors (for example, door status and environmental sensors). Huawei also provides theintelligent rPDU PDU2000M, whosemain control moduleincludes communications and DI interfaces designed to directly接入cabinet-level intelligent devices for centralized monitoring and management. In Huawei’s solution description for rPDU-monitoring networking, thePDU2000M is positioned to replace the expansion-module role, meaning that in applicable scenarios, cabinets can connect monitoring devices through the PDU2000M without deploying an additional UIM20A expansion module. This supports practical connections such asT/H sensors,smart U space managers, anddoor status/door-access related signalsthrough the PDU2000M interfaces, and then these devices are managed/bound in the controller’s smart module view. Huawei O&M guidance also emphasizes avoiding duplicate connections (do not connect the same device to both UIM20A and PDU2000M), aligning with the replacement concept. (Scribd)

In Huawei data center power distribution design, theSPD (Surge Protective Device)provides overvoltage surge protection for sensitive loads by diverting transient energy to ground through internal protection components. The SPD status indicator is used for O&M judgment of protection availability. When the indicator isgreen, it indicates the SPD protection components arewithin normal working conditionand the surge protection path remains effective. When the indicator turnsred, it typically means the SPD’s internal protection element hasdegraded or disconnected(end-of-life or failure after cumulative surges), and the device can no longer provide the rated protection level. In this condition, the distribution system is left exposed to surge risk, especially during thunderstorms, switching operations, or upstream disturbances. Therefore, Huawei O&M practice treats ared SPD indicator as a replacement trigger, requiring immediate corrective action: confirm alarm/status, ensure safe isolation following electrical safety procedures, and replace the SPD module to restore protection and maintain power system reliability.

In Huawei NetEco access design, the Web client is reached usingHTTPSto protect credentials and management data in transit. During initial deployment and commissioning, engineers access the NetEco login page by entering the server IP address followed by thedefault HTTPS service port. Huawei’s standard login method specifies that users open a browser and enterhttps://NetEco server IP:31943to access the NetEco login interface. This port is therefore treated as the default Web login port in typical on-premises NetEco deployments, and it is the baseline value referenced in installation and login guidance. Other ports may appear in specific tools, customized scenarios, version variants, reverse-proxy deployments, or when security policy requires port changes, but those are exceptions driven by planning or product packaging. For exam and commissioning purposes, the correct default login port is the one used by the NetEco Web client entry point under the standard deployment model:31943.

In UPS5000-H operation,normal modemeans the load is supplied by theinverter, and the rectifier/charger maintains the DC link and batteries. Themaintenance bypass switch(maintenance bypass breaker) is a mechanical isolation path intended to keep the load powered while the UPS is taken out of service for inspection or repair. Because it changes the power path and can create unsafe or unintended parallel paths if operated incorrectly, it isnot permitted to simply close the maintenance bypass switch while the UPS is in normal mode. The correct O&M logic is to perform acontrolled transferfirst (typically transferring the load tostatic bypasswith proper synchronization and confirmation that bypass power is healthy), and only then close the maintenance bypass switch so the UPS can be safely isolated. This prevents disturbance to the load, avoids synchronization/transfer hazards, and protects personnel from incorrect switching operations. Therefore, the statement is true in the sense of standard UPS5000-H operating procedure and safety interlock intent.

In Huawei data center facility management,NetEcois the management software platform (server/application) used to centrally monitor and manage sites, whileiBMCis the out-of-band hardware management interface used forserver remote maintenance(such as power control, hardware health, BIOS settings, and remote console) on the specific server where iBMC exists. TheiBMC port is not a service access interface for NetEco. NetEco’s backend services are accessed through theserver operating system network interfaces(management/service network) and corresponding application ports after the server is properly networked and configured. Even if NetEco is deployed on a Huawei server that has iBMC, iBMC only provides a channel to manage the server hardware and does not expose NetEco application access as an official commissioning/O&M path. Therefore, NetEco backend access should be planned via the site’s management network (IP addressing, routing, security policy), not via iBMC.

In Huawei UPS operating logic and standard data center O&M practice, themaintenance bypassis a mechanical path used to isolate the UPS for service. Innormal modethe load is carried by the inverter, so you shouldnot directly close the maintenance bypasswithout a controlled transfer procedure; doing so can create an unsafe parallel path or impact synchronization, therefore statementA is false. Instatic bypass mode, the load is supplied through the bypass path; if the bypass/mains source fails, the UPS can perform an automatic transfer from bypass to inverter, and the inverter will be supported by thebatterythrough the DC link, soB is trueunder the expected functional behavior.UPS5000-His designed as amodular UPS, supportingonline (hot) replacementof power modules to restore capacity without shutting down the entire system, soC is true. If therectifier mains input and bypass input share the same source, a mains outage removes the bypass source as well, meaning bypass operation is not available, soD is true.