Data Center

Trustworthy Data Center

Trustworthy Data Center

Data Center Security
Secure Location

Data centers are built in geographically secure locations where public facilities are within easy reach, locations generally safe from natural disasters and other risks. Different availability zones (AZs) are physically isolated from each other.

Reliable Architecture

Our full-stack data centers use a mature high reliability architecture that is always evolving. The architecture provides redundancy and protects upper-layer services from data center faults.

Close Monitoring

We have entire campuses dedicated to our data centers and use an access control system, a video surveillance system, and an infrared system to monitor and control the access to every cabinet unit, ensuring the data centers are physically safe.

 Personnel Management
Least Privileges Granted

HUAWEI CLOUD grants the minimum access possible to any personnel who need to enter or exit data centers. Before entering any data centers, the personnel must report what they plan to do there, pass strict reviews, and obtain approval. They can only enter allowed areas during approved and scheduled time periods, and their every operation in the data centers are monitored and regulated. Their access permissions will also expire after a predefined period of time.

Routine Permission Review

HUAWEI CLOUD routinely reviews access permissions to data centers and revokes unnecessary permissions to ensure your data is kept safe.

Asset Management

HUAWEI CLOUD data centers use an asset management system to track and record the locations, statuses, and other information about assets throughout their lifecycles. Storage media are strictly managed and controlled. At the end of their lifecycles, they will be degaussed to destroy data and prevent any possible data leakage.

Monitoring and Tracing
Intelligent Monitoring

HUAWEI CLOUD data centers use an intrusion detection system to detect and monitor security events. If an intrusion is detected, the system will immediately report an alarm and inform onsite security, helping them quickly respond to the event.

Evidence Tracing

In compliance with laws and regulations, HUAWEI CLOUD data centers monitor and record equipment rooms and other important areas through CCTV surveillance, and keep the surveillance video for a certain period of time for event tracing. The data centers are monitored 24/7 by both on-duty personnel and an automatic security system.

Design, Construction, and O&M

We strictly control our data center quality. Our design, construction, and O&M teams work together to develop and implement management regulations for the entire delivery process, including building design and review, equipment checks before acceptance, construction process control, acceptance unit tests, joint commissioning tests, quality checks and repairs before usage, and O&M and tool standardization. All these make sure the new data centers comply with the high quality design standards, so we have a solid foundation for reliability.

Preventive Infrastructure Maintenance
Infrastructure Maintenance and Repair

We have developed a maintenance management system (MMS) and annually formulate preventive maintenance (PM) plans (including forecast maintenance [FM] plans) and other O&M work plans, emergency operations plans (EOPs), Methods of Procedures (MOPs), Standard Operating Procedures (SOPs), and training plans to guide professional, regulation-compliant O&M. We ensure device stability by regularly overhauling equipment, performing routine general inspections, and through self-checks, and expert inspections.

Efficient O&M

Our experts regularly perform inspections across regions, evaluate data center trustworthiness, and help the data centers improve their work. We also keep improving O&M management processes through a Plan-Do-Check-Act (PDCA) process, always working to improve efficiency.

Infrastructure Service SLAs

We use a data center infrastructure management (DCIM) system and other visualized tool platforms to manage the schedule, workload, and progress of preventative O&M, satisfying customer Service Level Agreements (SLAs).

 

Facility Performance Monitoring
24/7 Power Supply

Our data centers use redundant power supplies managed by the DCIM system. If any device or system fault or loss of redundancy occurs, DCIM immediately reports an alarm and implements a contingency plan to troubleshoot the fault, ensuring the data center receives power 24/7.

Intelligent Cooling

Data center temperature is controlled by an intelligent cooling system to keep servers working at the most appropriate temperature. If any device/system fault occurs, DCIM reports an alarm within seconds and implements a contingency plan to troubleshoot the fault, ensuring operating temperatures are stable 24/7.

Fire Monitoring and Extinguishing

Our data centers use fire suppression systems that detect any early signs of a fire, automatically report alarms, and take measures to extinguish fires if they occur. Main fire protection actions and alarm information are managed in the DCIM system.  

Physical Security Monitoring  

We deploy monitoring systems and take control measures at four different levels of granularity: campuses, buildings, data equipment rooms, and equipment rooms and cabinets. We coordinate manpower, facilities, and technologies to physically protect our data centers.

Spare Parts Plan

To minimize downtime, data centers implement spare parts plans based on the importance, lifecycles, failure rates, procurement periods, and storage conditions of various device parts. Spare parts of the required quantity and specifications are immediately available in case of a device failure.

Business Continuity Plan (BCP)

The HUAWEI CLOUD data center BCP clearly defines the measures to prevent unexpected power-offs, cooling failures, security system failures, or fires. BCP provides contingency and verification plans. We have also developed processes for infrastructure event management, change management, risk management, and problem management.

Epidemic Prevention and Response

Epidemic prevention and response are an important part of our disaster recovery plan, which is constantly updated in practice. Each level of granularity, each area, and each role has its own personnel configurations and contingency plans to ensure there is adequate manpower to guarantee asset security and business continuity during an epidemic. All our epidemic prevention and response measures comply with the laws and regulations of local health institutions.

Quick Response

Data center personnel live near their workplaces and work in rotating shifts to provide 24/7 support. We have signed quick response contracts with our equipment vendors to ensure we can get their technical and spare parts services immediately when needed. Our fuel and water supply in emergencies are guaranteed by SLAs.

Emergency Management

We have a set of procedures and standards for infrastructure emergency management in a range of categories, including personal safety assurance, infrastructure, natural disasters, and public health. Each category has an EOP that is regularly updated after drills and reviews. These guides help data center personnel in quick response and troubleshooting, ensure infrastructure reliability, reduce the duration of redundancy loss, and shorten the mean time to repair (MTTR).