Data center equipment putaway method and system based on dynamic load balancing

Abstract

The invention discloses a data center equipment racking method and system based on dynamic load balancing, and the method comprises the steps: S1, obtaining the resource information of cabinets, and determining the available space and available power of each cabinet; when the cabinet is an alternating current cabinet, the cabinet resource information further comprises a three-phase unbalance degree; s2, acquiring a required U bit and power of equipment to be loaded into a cabinet, and generating a candidate cabinet list meeting the loading requirement based on the cabinet resource information; s3, setting a first weight parameter and a second weight parameter for the direct current cabinet and the alternating current cabinet respectively, and performing comprehensive scoring on the candidate cabinet list according to a preset scoring weight; for the direct current cabinet, according to the available space and the available electric quantity, carrying out weighted summation by combining a preset first weight parameter; for the AC cabinet, according to the available space, the available electric quantity and the three-phase unbalance degree, performing weight summation by combining a preset second weight parameter; and S4, forming a descending recommendation sequence according to a weighted summation result, and recommending an optimal cabinet.

Description

Technical Field

[0001] This invention relates to the field of data center operation and maintenance management technology, and more specifically to a method and system for mounting data center equipment based on dynamic load balancing. Background Technology

[0002] With the rapid development of technologies such as cloud computing, big data, and artificial intelligence, data centers, as the core infrastructure of the digital economy, are experiencing a continuous increase in scale, density, and complexity. The number of IT devices such as servers, network equipment, and storage devices within data center rooms is growing exponentially, making efficient and rational equipment deployment a crucial aspect of data center operation and maintenance management. Currently, mainstream data center equipment deployment strategies primarily rely on manual planning and experience-based judgment, or are assisted by semi-automated tools based on simple rules. Operation and maintenance personnel manually select racks and storage units based on basic information such as equipment type, power, and network connection requirements, combined with the data center's usage, relying on personal experience. This approach is inefficient, susceptible to subjective factors, and struggles to cope with the complexity of large-scale, high-density data centers. Some automated tools employ preset, relatively fixed rules, such as "centralized racking by brand / model," "regional division by business system," and "filling racks first, then deploying." While these rules can improve efficiency to some extent, they often lack a holistic optimization perspective, failing to simultaneously and dynamically consider multiple interdependent key factors, and frequently neglecting power load balancing. Existing strategies are usually planned based on the current static snapshot of the data center, which makes it difficult to effectively incorporate dynamic data during the operation of the data center (such as real-time power consumption, temperature distribution, and three-phase imbalance). This leads to frequent local hot spots, overloads, and power quality problems after the racks are put into use, and the planning results are out of touch with the actual operation requirements.

[0003] In summary, traditional or existing data center equipment racking strategies generally suffer from problems such as low planning efficiency, limited optimization dimensions, reliance on manual experience, and a lack of global and dynamic optimization capabilities. This results in the inefficient utilization of data center resources such as space, power, and cooling, increases operating costs, creates security risks, and reduces the agility of operation and maintenance management.

[0004] Therefore, how to improve the efficiency of product listing and optimize resource utilization is an urgent problem to be solved in this field. Summary of the Invention

[0005] In view of the above problems, the present invention is proposed to provide a method and system for mounting data center equipment based on dynamic load balancing to overcome or at least partially solve the above problems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for mounting data center equipment based on dynamic load balancing, characterized by the following steps: S1: Obtain rack resource information in real time to confirm the available space and available power of each rack; when the rack is an AC rack, the rack resource information also includes three-phase imbalance. S2: Obtain the required U-position and power of the equipment to be racked, and dynamically generate a list of candidate racks that meet the racking requirements based on the rack resource information; S3: Set a first weight parameter and a second weight parameter for DC and AC cabinets respectively, and give a comprehensive score to the candidate cabinet list according to the preset scoring weights; For DC cabinets, a weighted sum is performed based on the available space and available power, combined with a preset first weighting parameter; For the AC cabinet, a weighted sum is performed based on the available space, the available power, and the three-phase imbalance, combined with a preset second weighting parameter. S4: Based on the weighted summation result, a descending recommendation sequence is generated, and the best server rack is recommended.

[0007] Preferably, the steps further include: S5: Identify the rack and update the rack resource information in real time based on the used resource information after racking, realizing a real-time decision-making closed loop of "racking and adjusting at the same time" during the equipment racking process.

[0008] Preferably, obtaining cabinet resource information in S1 includes: establishing communication with the network device management system and / or the environmental monitoring system to obtain the cabinet resource information in real time.

[0009] Preferably, the racking decision adopts a multi-factor weighting model: the three factors of power load, space occupied and power balance are quantified into scores of 0-100, and the "racking priority score" is dynamically generated by weighted summation. The rack with the highest score wins. After each racking, the weight is recalculated and the "best rack sequence" of subsequent racking devices is rearranged in real time.

[0010] Preferably, the steps further include: after obtaining the rack resource information, performing data preprocessing, the preprocessing including: Identify and filter outliers, null values, and out-of-range data; The collected data is standardized, including: unit normalization, caliber completion, U-position resolution, precision alignment, and confidence level labeling.

[0011] Preferably, the comprehensive scoring steps include: The available space items are confirmed based on the available space. Confirm the available power quantity based on the available power quantity; The three-phase imbalance term is identified based on the aforementioned three-phase imbalance degree.

[0012] A data center equipment racking system based on dynamic load balancing, comprising: The data acquisition module is used to acquire rack resource information and confirm the available space and available power of each rack; when the rack is an AC rack, the rack resource information also includes three-phase imbalance. The rack selection module is used to obtain the required U-position and power of the equipment to be racked, and generate a candidate rack list that meets the racking requirements based on the rack resource information; it is used to comprehensively score the candidate rack list according to preset scoring weights; for DC racks, it performs a weighted sum based on the available space and available power, combined with a preset first weight parameter; for AC racks, it performs a weighted sum based on the available space, available power, and three-phase imbalance, combined with a preset second weight parameter. The recommendation module is used to generate a descending recommendation sequence based on the weighted summation result and recommend the best server rack.

[0013] Preferably, the data acquisition module is provided with a communication interface, which is used to connect to an external network device management system and / or environmental monitoring system.

[0014] Preferably, it also includes a visual human-computer interaction module, the visual human-computer interaction module comprising: The strategy configuration submodule is used to set the first weight parameter and the second weight parameter for DC and AC cabinets respectively. The weight coefficients can be dynamically shifted according to the season or business peak and valley. The device racking module is used to display the resource usage of each rack and the sorting of the recommended sequence, which is used to select a rack for racking.

[0015] As can be seen from the above technical solution, compared with the prior art, this invention discloses a method and system for data center equipment racking based on dynamic load balancing. By collecting real-time data on available space and power supply for each rack, as well as the three-phase imbalance of AC racks, candidate racks capable of supporting the equipment to be racked are first selected. Then, differentiated weight parameters are set for DC or AC racks, and a weighted score is applied to the space, power load, and / or the three-phase imbalance of AC racks. Rational racks with higher scores and more reasonable load redundancy are prioritized, avoiding single-rack overload or resource idleness. Simultaneously, rack resource information is updated in real-time after racking to ensure that subsequent decisions are based on the latest load status, from local rack load adaptation to global resource allocation, achieving dynamic load balancing for data center equipment racking. Through automated data collection, processing, and recommendation, manual intervention is significantly reduced, improving racking efficiency. The comprehensive weighting considers space, power, and power quality to achieve balanced resource utilization and avoid local overload. DC racks are exempt from imbalance items, while AC racks introduce imbalance, achieving fair scoring. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of a data center equipment racking method based on dynamic load balancing provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of a data center equipment racking system based on dynamic load balancing provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figure 1 This invention discloses a method for mounting data center equipment based on dynamic load balancing, comprising the following steps: S1: Obtain rack resource information and confirm the available space and available power of each rack; when the rack is an AC rack, the rack resource information also includes three-phase imbalance. S2: Obtain the required U-position and power of the equipment to be racked, and generate a list of candidate racks that meet the racking requirements based on rack resource information; S3: Set the first weight parameter and the second weight parameter for DC and AC cabinets respectively, and give a comprehensive score to the candidate cabinet list according to the preset scoring weights; For DC cabinets, a weighted sum is performed based on available space and available power, combined with a preset first weighting parameter; For AC cabinets, a weighted sum is performed based on available space, available power, and three-phase imbalance, combined with a preset second weighting parameter. S4: Based on the weighted summation result, a descending recommendation sequence is generated, and the best server rack is recommended.

[0020] To further implement the above technical solution, the steps also include: S5: Identify the rack and update the rack resource information according to the used resource information after racking.

[0021] In this embodiment, after the device is put on the rack, the information of the used resources is written back in real time. When a new device needs to be put on the rack next time, the data obtained by this racking method is always based on the latest snapshot.

[0022] To further implement the above technical solution, S1 involves obtaining rack resource information, including: establishing communication with the network equipment management system and / or the environmental monitoring system to obtain the rack resource information in real time.

[0023] Specifically, the system uses the Network Equipment Management System (NEMS) and Environmental Monitoring System (EMS) interfaces to capture real-time data on the total number of storage units (U-bits), the number of used U-bits, the rated power (kW), and the power consumption (kW) of each cabinet. For AC rack-mount cabinets, it collects the real-time current of the A, B, and C phases and marks the cabinet's power supply type as DC / AC / AC-DC hybrid. During write-back, data can be directly written back to the Network Equipment Management System (NEMS) and the Environmental Monitoring System (EMS).

[0024] To further implement the above technical solution, the steps also include: after obtaining the rack resource information, performing data preprocessing, which includes: identifying and filtering outliers, null values ​​and out-of-range data; and standardizing the collected data, including: unit normalization, caliber completion, U-bit resolution, precision alignment and confidence labeling.

[0025] Specifically, for unit normalization, numbers and units can be captured using regular expressions, and then uniformly converted according to the units, with numbers standardized to a specific format, such as retaining 3 decimal places; for U-bit parsing, the space of auxiliary equipment can be removed from the basic space to confirm the valid space; for confidence labeling, corresponding weights can be set according to the credibility of the data source, such as a confidence level of 1.0 for a source of nameplates or SNMP read-only fields; a confidence level of 0.8 for a source of manual Excel; and a confidence level of 0.7 if converted / estimated.

[0026] Furthermore, by capturing information such as the total number of USB ports, the number of USB ports used, the rated power (kW), and the power consumption (kW) of each cabinet in real time through the interfaces of the Network Equipment Management System (NEMS) and the Environmental Monitoring System (EMS), the available power, available space, and corresponding utilization rate can be calculated. For AC cabinets, the three-phase imbalance can be calculated.

[0027] To further implement the above technical solution, the comprehensive evaluation steps include: The available space item is determined based on the available space; for example, the maximum available space is 100 points, and other racks are calculated as available space / maximum available space * 100.

[0028] The available power level is determined based on the available power capacity; for example, the highest available power capacity is scored as 100 points. Other racks are scored as available power capacity / maximum available power capacity * 100.

[0029] The three-phase imbalance item is determined based on the three-phase imbalance degree; the lowest three-phase imbalance degree is 100 points. For other cabinets, the score is calculated as 100 - (three-phase imbalance degree / maximum three-phase imbalance degree) * 100.

[0030] In this embodiment, each cabinet is traversed according to the candidate list, and the type label of each cabinet is identified. For DC cabinets and AC cabinets, the corresponding scoring items are combined with preset weight coefficients to perform weighted summation to obtain the final score.

[0031] like Figure 2 Based on the same inventive concept, this invention discloses a data center equipment racking system based on dynamic load balancing, comprising: The data acquisition module is used to obtain rack resource information and confirm the available space and available power of each rack; when the rack is an AC rack, the rack resource information also includes three-phase imbalance. The rack selection module is used to obtain the required U-position and power of the equipment to be racked, and generate a list of candidate racks that meet the racking requirements based on rack resource information; it is used to comprehensively score the candidate rack list according to preset scoring weights; for DC racks, it performs a weighted sum based on available space and available power, combined with a preset first weight parameter; for AC racks, it performs a weighted sum based on available space, available power, and three-phase imbalance, combined with a preset second weight parameter. The recommendation module is used to generate a descending recommendation sequence based on the weighted summation result and recommend the best server rack.

[0032] To further implement the above technical solution, a visual human-computer interaction module is included. The visual human-computer interaction module includes: a strategy configuration submodule, used to set a first weight parameter and a second weight parameter for DC and AC cabinets respectively; and a device racking submodule, used to display the resource occupancy of each cabinet and the sorting of the recommended sequence, for the selected cabinet to perform racking.

[0033] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0034] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for mounting data center equipment based on dynamic load balancing, characterized in that, The method comprises the following steps: S1: obtaining cabinet resource information, confirming the available space and available power of each cabinet; when the cabinet is an AC cabinet, the cabinet resource information further comprises a three-phase imbalance degree; S2: obtaining the demand U-bit and power of the equipment to be placed in the cabinet, and generating a candidate cabinet list meeting the placement demand based on the cabinet resource information; S3: setting a first weight parameter and a second weight parameter for DC and AC cabinets respectively, and comprehensively scoring the candidate cabinet list according to a preset scoring weight; For a DC cabinet, the available space and available power are combined with the preset first weight parameter for weighted summation; For an AC cabinet, the available space, available power and three-phase imbalance degree are combined with the preset second weight parameter for weighted summation; S4: forming a descending recommendation sequence according to the weighted summation result, and recommending the best cabinet.

2. The data center equipment on-rack method based on dynamic load balancing of claim 1, wherein, The steps further comprise: S5: identifying the rack, and updating the cabinet resource information according to the used resource information after the rack is placed.

3. The method of claim 1, wherein, In S1, the cabinet resource information is obtained by establishing communication with a network device management system and / or a dynamic environment monitoring system, and the cabinet resource information is obtained in real time.

4. The method of claim 1, wherein, The steps further comprise: after the cabinet resource information is obtained, data preprocessing is performed, and the preprocessing comprises: identifying and filtering abnormal values, null values and out-of-range data; standardizing the collected data, including unit normalization, caliber completion, U-bit analysis, precision alignment and confidence marking.

5. The method of claim 1, wherein, The comprehensive scoring step comprises: confirming the available space item according to the available space; confirming the available power item according to the available power; confirming the three-phase imbalance item according to the three-phase imbalance degree.

6. A dynamic load balancing based data center equipment racking system, comprising: It comprises: a data acquisition module for obtaining cabinet resource information and confirming the available space and available power of each cabinet; when the cabinet is an AC cabinet, the cabinet resource information further comprises a three-phase imbalance degree; a cabinet screening module for obtaining the demand U-bit and power of the equipment to be placed in the cabinet, and generating a candidate cabinet list meeting the placement demand based on the cabinet resource information; and for comprehensively scoring the candidate cabinet list according to a preset scoring weight; for a DC cabinet, the available space and available power are combined with the preset first weight parameter for weighted summation; for an AC cabinet, the available space, available power and three-phase imbalance degree are combined with the preset second weight parameter for weighted summation; a recommendation module for forming a descending recommendation sequence according to the weighted summation result, and recommending the best cabinet.

7. The dynamic load balancing based data center equipment racking system of claim 6, wherein, The data acquisition module is provided with a communication interface for connecting an external network device management system and / or a dynamic environment monitoring system.

8. The data center equipment on-rack system based on dynamic load balancing according to claim 6, wherein, It further comprises a visual human-computer interaction module, which comprises: a strategy configuration sub-module for setting a first weight parameter and a second weight parameter for DC and AC cabinets respectively; a device placement sub-module for displaying the resource occupation of each cabinet and the sorting of the recommendation sequence, and for selecting the cabinet for placement.

Images