Huawei H20-923_V1.0 - HCSP-Field-Data Center Facility V1.0

NetEco is an enterprise management application with specific runtime dependencies, including supported operating system versions, database/middleware compatibility, CPU/memory/storage requirements, and network/security prerequisites. Because of these dependencies, Huawei does not treat NetEco as “hypervisor-agnostic for any virtualization platform.” In practice, NetEco deployment in virtualized environments is supported only when the virtualization stack meets Huawei’s compatibility requirements, such as validated hypervisors/virtualization solutions, proper virtual hardware settings, and resource reservation rules to ensure stable performance (especially for alarm processing, historical data storage, and northbound integration tasks). If deployed on an unverified virtualization platform, issues may occur such as clock drift, I/O latency, database performance bottlenecks, unstable NIC behavior, or unsupported VM device models, which can impact data collection, event correlation, and overall system stability. Therefore, the correct interpretation is that NetEco can be deployed onsupported/verifiedvirtualization platforms, not onanyvirtualization platform.

In Huawei NetEco statistical reporting, the query range is constrained by the report data retention policy and how summary data is stored. Forhourly statistics, the platform keeps a limited window of high-granularity data to balance storage consumption and performance. The hourly report retention is120 hours, which equals5 days. That is why hourly statistics queries are limited to5 days—beyond that window, hourly granularity is no longer retained in the report database for direct querying. For longer time horizons, NetEco relies on lower-granularity summaries (daily/weekly/monthly/yearly). Formonthly statistics, the platform supports long-term trend analysis and KPI tracking, and the maximum historical horizon aligns with the long-term statistics retention boundary used by NetEco reporting, which isup to 5 years. This design ensures fast query response, manageable database growth, and clear separation between short-term operational detail (hourly) and long-term management analysis (monthly/yearly).

After ECO mode is enabled on the UPS5000-H, the system’s control objective changes from “always supply the load through the inverter” to “maximize efficiency while keeping protection mechanisms ready.” When themains input is within the configured ECO acceptance window(voltage, frequency, THD, and other quality thresholds), the UPS supplies the critical load primarily through thestatic bypass path, which has lower conversion losses than continuous double-conversion. The inverter remains synchronized and on standby, so if the mains quality deteriorates or an abnormal condition is detected, the UPS canrapidly transferthe load back to inverter operation to maintain output stability. Battery mode is only used when the rectifier/bypass source is unavailable or out of range, and maintenance bypass is a manual service path used for isolation during maintenance, not for normal energy-saving operation. Therefore, with stable utility power and ECO configured, the UPS5000-H mainly operates instatic bypass mode.

Wizard startup commissioning is designed to verify that the unit’skey controllable subsystemscan be started, regulated, and protected correctly. Conditions that prevent a subsystem from operating normally will directly cause commissioning failure. If thecondensate pump is stuck, the drainage function cannot be validated and water may accumulate in the drain pan, which is treated as a functional fault during commissioning. Ifindoor fan 2 drive is faulty, the fan system cannot meet airflow requirements or redundancy expectations, so the fan commissioning item fails. If theelectronic expansion valve (EEV) is not opened, refrigerant flow and cooling control cannot be established, so the cooling system commissioning fails because the unit cannot build a stable refrigeration cycle or reach expected operating parameters.

In contrast, thewater leakage ropeis aprotective detection accessoryused for leak sensing and alarming. While it is important for site safety and recommended for operation, its absence typically does not block the unit from completing the functional commissioning steps for fans, cooling, and drainage; it mainly affects leak detection coverage and related alarms rather than the basic startup commissioning pass/fail.

This statement is true because post-installation verification is a mandatory acceptance step for battery systems in data centers. After a battery module is installed,torque verificationensures that all power terminals, interconnect bars, and grounding points meet the specified tightening requirements. If torque is too low, contact resistance increases, which can cause abnormal heating, voltage drop, nuisance alarms, and even connector damage during high current charge/discharge. If torque is too high, threads, lugs, or busbars can be damaged, creating hidden reliability risks.Insulation testingconfirms that the battery circuit has no unintended leakage paths to chassis or ground and that cable routing, connector seating, and protective covers are correct. This is especially important for lithium battery cabinets where high DC voltage and large fault current capability demand strict electrical isolation. Performing torque and insulation checks after installation reduces early-life faults, improves safety, and ensures the cabinet can pass commissioning tests and enter stable service under UPS charge/discharge control.

In Huawei precision cooling units, the LCD controller (human–machine interface) must exchange real-time operating data and commands with the unit’s main control module while also receiving a stable low-voltage power supply. This connection is typically implemented as alow-voltage communication bus plus DC power, which is why the interface is identified asI2C/12V. TheI2Cportion provides the data channel used for status display, parameter viewing/setting (within permission scope), alarm presentation, and command delivery (such as mute, reset, or control enable actions). The12 Vportion supplies power to the controller so it can operate independently of high-voltage power circuits and remain safe to access during routine O&M. Interfaces likeDPandUSBare not used as the internal control link between the HMI and the main control board in this type of unit, and “SW” is not the designated controller-to-main-board interface in the standard wiring definition. Therefore,I2C/12Vis the correct selection.

Huawei data center power distribution design prioritizes continuity for critical loads. For single-cord IT equipment that cannot accept dual inputs, an STS provides a practical high-availability approach by connecting the load to two independent AC sources (often A and B feeds) and transferring the load to the healthier source when abnormalities occur. The key value is speed and coordination: the STS detects voltage/frequency deviations or source failure and performs a rapid transfer to minimize or eliminate interruption seen by the load. This supports O&M strategies where upstream systems (UPS, PDUs, distribution boards) are maintained or isolated without forcing downtime on single-cord devices. In routine operations, STS status, transfer logs, alarms, and source quality parameters are monitored to prevent nuisance transfers and to ensure correct upstream selectivity and grounding. Correct commissioning also ensures the two sources meet synchronization conditions and that bypass/maintenance procedures are standardized, improving safety and reducing operational risk.

ECO mode is designed to improve UPS efficiency by supplying the load mainly through the static bypass path when the mains is within the acceptable quality window, while keeping the inverter synchronized and ready to take over quickly if power quality degrades. However, when the load istoo light, the overall operating characteristics of the UPS system are less favorable for ECO operation. At low load levels, the benefits of ECO in reducing conversion losses are smaller, and the system becomes more sensitive to frequent transfers caused by minor mains fluctuations, which can increase transfer次数and operational risk. In addition, very light loads can reduce the effectiveness of system self-check behaviors and may lead to less stable operating margins for parallel systems, power modules, and downstream distribution, depending on site conditions. Therefore, Huawei O&M guidance sets a practical threshold and states thatECO mode is not recommended when the load is lower than 20%, helping ensure stable operation, fewer unnecessary transfers, and better overall power supply reliability for critical loads.