Distributed node control method, device and equipment based on differentiated quota scheduling

Abstract

Embodiments of the present disclosure disclose a distributed node control method, device and equipment based on differentiated quota scheduling. A specific implementation of the method comprises: obtaining a global quota upper limit corresponding to a node running information set and a distributed node set; performing elastic quota division processing on the distributed node set to obtain an initial quota information set; performing over-limit convergence processing on the initial quota information set according to the global quota upper limit to obtain a quota-in-constraint information set; performing occupation trend identification processing on the quota-in-constraint information set to obtain a supply node set; performing steady-state redistribution processing on the quota-in-constraint information set to obtain a target quota information set; performing change determination processing on the distributed node set to obtain a quota change information set; and performing batch load control on the distributed node set. The implementation can reduce the system resource consumption and load of the distributed node and improve the running stability of the distributed node.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and more specifically to a distributed node control method, apparatus, and device based on differentiated quota scheduling. Background Technology

[0002] Distributed node control refers to a technology that sets a transmission rate limit for each of multiple nodes sharing the same total transmission limit and applies control accordingly. In securities trading, a securities company's various securities trading operations are carried out by independent order placement systems. Each order placement system is deployed in the company's data center and sends orders to the exchange's host through the same order placement gateway. The exchange controls the total order flow rate through this order placement gateway. When the actual order flow rate reaches the flow rate limit, throttling is initiated, and the excess is postponed to the next time slot. This flow rate limit can only be an integer multiple of the single-order flow rate limit stipulated by the exchange and cannot be changed within a trading day. However, the order placement rates of each order placement system are not stable. The market-making system must continuously provide two-sided quotes, and its order placement rate remains continuous during continuous bidding periods. The order placement rates of other order placement systems surge during market fluctuations. Therefore, securities companies have a high demand for setting differentiated transmission rates based on the differences and changes in the actual usage of each order placement system, within the flow rate limit of the order placement gateway.

[0003] Currently, the common approach for distributed node control based on differentiated quota scheduling is to collect the actual transmission volume of each distributed node within a unit of time according to the scheduling cycle, redetermine the quota threshold of each node according to a pre-set fixed ratio or according to the proportion of the actual transmission volume of each node in the current cycle in the total, and limit the request transmission rate of each node with the redetermined quota threshold.

[0004] However, in practice, it has been found that when using the above method to control distributed nodes, the following technical problems often arise: Since the quota thresholds for each node are uniformly redefined according to a fixed ratio or the single-point transmission volume of the current period, the transmission rate of some nodes is continuously limited, resulting in request backlog, while the quota thresholds of other nodes are consistently higher than the actual transmission volume. Furthermore, when the transmission volume of some nodes is at the starting point of an increase, the above method still lowers the quota thresholds of these nodes based on the current remaining amount, causing these nodes to reach the lowered thresholds in subsequent periods. Simultaneously, the quota thresholds increase and decrease between nodes with higher and lower current occupancy, and both surge simultaneously in subsequent periods. The allocated quotas are reversed in the next scheduling cycle, causing a continuous mismatch between the quota thresholds and the actual transmission volume of each node. This leads to increased system resource consumption, higher node load, and reduced operational stability of the distributed nodes.

[0005] The information disclosed in this background section is only intended to enhance the understanding of the background of the present disclosure concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0007] Some embodiments of this disclosure propose a distributed node control method, apparatus, and device based on differentiated quota scheduling to solve one or more of the technical problems mentioned in the background section above.

[0008] In a first aspect, some embodiments of this disclosure provide a distributed node control method based on differentiated quota scheduling, comprising: in response to receiving a quota scheduling request for a distributed node set, obtaining a node operation information set and a global quota upper limit corresponding to the distributed node set; performing elastic quota allocation processing on the distributed node set according to the node operation information set to obtain an initial quota information set; in response to determining that the initial quota information set meets a preset over-limit condition relative to the global quota upper limit, performing over-limit convergence processing on the initial quota information set according to the global quota upper limit to obtain a constrained quota information set; performing occupancy trend identification processing on the constrained quota information set to obtain a supply node set; performing steady-state redistribution processing on the constrained quota information set according to the supply node set to obtain a target quota information set; performing change determination processing on the distributed node set according to the target quota information set and the initial quota information set to obtain a quota change information set; and performing batch load control on the distributed node set according to the quota change information set and the target quota information set.

[0009] Secondly, some embodiments of this disclosure provide a distributed node control device based on differentiated quota scheduling, comprising: an acquisition unit configured to, in response to receiving a quota scheduling request for a distributed node set, acquire a node operation information set and a global quota upper limit corresponding to the distributed node set; a partitioning unit configured to, based on the node operation information set, perform elastic quota partitioning processing on the distributed node set to obtain an initial quota information set; and a convergence unit configured to, in response to determining that the initial quota information set satisfies a preset over-limit condition relative to the global quota upper limit, perform a convergence process on the initial quota information set according to the global quota upper limit. The system performs over-limit convergence processing to obtain a quota information set within constraints; an identification unit is configured to perform occupancy trend identification processing on the quota information set within constraints to obtain a supply node set; a redistribution unit is configured to perform steady-state redistribution processing on the quota information set within constraints based on the supply node set to obtain a target quota information set; a determination unit is configured to perform change determination processing on the distributed node set based on the target quota information set and the initial quota information set to obtain a quota change information set; and a control unit is configured to perform batch load control on the distributed node set based on the quota change information set and the target quota information set.

[0010] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0011] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0012] The above embodiments of this disclosure have the following beneficial effects: The distributed node control method based on differentiated quota scheduling in some embodiments of this disclosure can reduce the system resource consumption of distributed nodes, reduce node load, and improve operational stability. Specifically, the reason for the increased system resource consumption, increased node load, and decreased operational stability of distributed nodes is that the quota threshold of each node is uniformly re-determined according to a fixed proportion or the single-point transmission volume of the current period. The transmission rate of some nodes is continuously limited, resulting in request backlog, while the quota threshold of other nodes is continuously higher than the actual transmission volume, causing a continuous mismatch between the quota threshold and the actual transmission volume of each node, leading to increased system resource consumption, increased node load, and decreased operational stability of distributed nodes. Based on this, the distributed node control method based on differentiated quota scheduling in some embodiments of this disclosure can first, in response to receiving a quota scheduling request for a distributed node set, obtain the node operation information set and the global quota upper limit corresponding to the distributed node set. Here, the aforementioned node operation information set provides the quota threshold, measured occupancy, and occupancy sequence for each distributed node. The aforementioned global quota upper limit provides the constraints on the sum of each quota threshold, providing a data foundation for subsequent allocation of cedable quota threshold reserves and determination of over-limit. Next, based on the aforementioned node operation information set, the aforementioned distributed node set undergoes elastic quota allocation processing to obtain an initial quota information set. Here, cedable quota threshold reserves are allocated separately for ordinary nodes, anchor nodes, and designated nodes, ensuring that the quota occupied by nodes that must retain basic sending capacity and nodes whose quota thresholds are not modified are not taken as cedable amounts. The resulting node elastic quota can quantify the actual cedable reserves that each distributed node can give up in this quota scheduling. Secondly, in response to determining that the aforementioned initial quota information set meets the preset over-limit conditions relative to the aforementioned global quota upper limit, the aforementioned initial quota information set undergoes over-limit convergence processing based on the aforementioned global quota upper limit to obtain a constrained quota information set. Here, ordinary nodes with larger quota threshold margins are prioritized to relinquish their margins according to tiers, bringing the sum of all quota thresholds back below the aforementioned global quota upper limit, while simultaneously reducing the number of nodes whose quota thresholds are overwritten. Then, the quota information set within the constraints is processed for occupancy trend identification to obtain the supply node set. Here, distributed nodes whose occupancy is at the starting point of an increase are excluded from the nodes that can relinquish their margins, preventing their quota thresholds from being reached in subsequent periods; simultaneously, distributed nodes whose occupancy has been increasing for a long time and whose rate of change has slowed down are retained among the nodes that can relinquish their margins, preventing the available margins from shrinking excessively due to exclusion. Afterwards, based on the aforementioned supply node set, the quota information set within the constraints is processed for steady-state redistribution to obtain the target quota information set.Here, nodes that relinquish their quotas retain a certain quota threshold margin after relinquishing them, while nodes receiving quotas have their quota threshold margins replenished, ensuring that the quota thresholds match the actual transmission volume of each node and reducing the likelihood of allocated quotas being reversed in the next scheduling cycle. Subsequently, based on the aforementioned target quota information set and the aforementioned initial quota information set, the aforementioned distributed node set undergoes change determination processing to obtain a quota change information set. Here, the quota thresholds before and after scheduling are compared node by node, ensuring that distributed nodes whose quota thresholds have not changed do not enter subsequent load control, reducing the number of nodes whose local rate limiting parameters and concurrency parameters are reset. Finally, based on the aforementioned quota change information set and the aforementioned target quota information set, the aforementioned distributed node set undergoes batch load control. Here, load control is first applied to distributed nodes whose quota thresholds have been lowered, and then to distributed nodes whose quota thresholds have been raised, ensuring that the sum of all quota thresholds never exceeds the aforementioned global quota upper limit throughout the entire load control process. Therefore, this distributed node control method based on differentiated quota scheduling can match the quota threshold with the actual transmission volume of each node, thereby reducing the system resource consumption of distributed nodes, reducing node load, and improving operational stability. Attached Figure Description

[0013] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0014] Figure 1 This is a flowchart of some embodiments of the distributed node control method based on differentiated quota scheduling according to this disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the distributed node control device based on differentiated quota scheduling according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0015] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0016] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0017] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0018] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0019] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0020] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] Figure 1 A flow 100 of some embodiments of a distributed node control method based on differentiated quota scheduling according to this disclosure is shown. This distributed node control method based on differentiated quota scheduling includes the following steps: Step 101: In response to receiving a quota scheduling request for the distributed node set, obtain the node running information set and the global quota limit corresponding to the distributed node set.

[0022] In some embodiments, the executing entity (e.g., an electronic device) of the above-described distributed node control method based on differentiated quota scheduling can, in response to receiving a quota scheduling request for a distributed node set, obtain the node operation information set and the global quota limit corresponding to the distributed node set via a wired or wireless connection. The distributed nodes in the distributed node set can be independently deployed, data-isolated devices that send requests to external hosts through a shared egress and are jointly constrained by the same total sending limit. The shared egress can be a gateway device used by all distributed nodes in the distributed node set to send requests to external hosts. For example, the distributed nodes can be server hosts hosting a securities trading system, with multiple server hosts sending orders to the exchange host through the same egress gateway. The quota scheduling request can be a request message used to trigger a re-determination of the quota threshold for each distributed node in the distributed node set. The quota scheduling request can be triggered periodically according to a preset scheduling period. The preset scheduling period can be a pre-set time interval between two adjacent quota scheduling events. For example, the preset scheduling period can be 24 hours. The node operation information in the aforementioned node operation information set can reflect the quota configuration and request sending status of distributed nodes at the current quota scheduling time. This node operation information may include quota thresholds, measured occupancy, minimum guaranteed value, node role identifier, and occupancy sequence. The quota threshold can be the upper limit of the number of requests a distributed node is allowed to send per unit time. The unit time can be a pre-set time length used to determine the request sending rate. For example, the unit time can be 1 second. The measured occupancy can be the number of requests actually sent by a distributed node per unit time at the most recent sampling position. The occupancy sequence can be a sequence obtained by arranging the measured occupancy of a distributed node at the most recent preset number of consecutive sampling positions in chronological order. The sampling position can be the time point on the time axis where the measured occupancy is collected, divided according to a preset sampling interval. The preset sampling interval can be a pre-set time interval between two adjacent sampling positions, and the preset sampling interval is less than the preset scheduling period. The aforementioned preset sampling quantity can be a pre-defined number of sampling locations. The preset sampling quantity and the preset sampling interval together determine the time range covered by the aforementioned occupancy sequence. For example, the preset sampling quantity can be 12. The last element of the aforementioned occupancy sequence represents the measured occupancy. The occupancy sequences corresponding to each distributed node in the aforementioned distributed node set correspond to the same set of sampling locations. The aforementioned minimum guarantee value can be the lower limit value that the quota threshold of a distributed node must not fall below in quota scheduling. The aforementioned node role identifier can be an identifier used to characterize the category to which the distributed node belongs in quota scheduling.The aforementioned global quota limit can be the maximum value that the sum of the quota thresholds of all distributed nodes in the aforementioned distributed node set must not exceed. This global quota limit is enforced by the aforementioned shared exit point. When the sum of the number of requests actually sent by each distributed node within a unit of time reaches the aforementioned global quota limit, the excess portion is postponed to the next time period. For example, the aforementioned global quota limit can be the flow rate limit value of the aforementioned reporting gateway. In practice, the aforementioned execution entity can first use the indicator tracking components (e.g., Micrometer, Prometheus client library) on each distributed node to accumulate and count the number of requests sent by the distributed nodes, obtaining an accumulated count value. Then, using a time-series database (e.g., Prometheus, VictoriaMetrics), the accumulated count value is retrieved from the indicator collection interface provided by the aforementioned indicator tracking components at the aforementioned preset sampling interval. Next, using the preset sampling interval as the sampling step size and the nearest preset number of sampling positions as the query interval, a range query is initiated on the cumulative count value using the rate function of the time series database (e.g., the rate function of Prometheus) to obtain the measured occupancy of each distributed node at the same set of sampling positions. Then, the measured occupancy at sampling positions with no sampling value in the range query results is set to 0. Subsequently, the measured occupancy of each distributed node at the same set of sampling positions is arranged in chronological order of the sampling positions to obtain an occupancy sequence. Finally, through the configuration pull interface of the configuration center (e.g., Nacos, Apollo), the quota threshold, minimum guarantee limit, and node role identifier of each distributed node, as well as the global quota limit corresponding to the distributed node set, are obtained.

[0023] Step 102: Based on the node operation information set, perform elastic quota allocation processing on the distributed node set to obtain the initial quota information set.

[0024] In some embodiments, the execution entity can perform elastic quota allocation processing on the distributed node set based on the node operation information set to obtain an initial quota information set. The initial quota information in the initial quota information set may reflect the quota threshold currently executed by the distributed nodes and its available vacancy. The node elastic quota may be the quota threshold vacancy that the distributed nodes can vacate in this quota scheduling. The node elastic quota may be a negative value. The quota threshold included in the initial quota information remains the same as the quota threshold included in the node operation information in this quota scheduling.

[0025] In some optional implementations of certain embodiments, the process of elastically allocating quotas to the distributed node set based on the node runtime information set to obtain an initial quota information set may include the following steps: The first step is to classify the distributed node set according to the node role identifier set included in the node operation information set, resulting in a set of ordinary nodes, a set of anchor nodes, and a set of designated nodes. Ordinary nodes in the ordinary node set can be distributed nodes that fully participate in both quota threshold increases and decreases during quota scheduling. Anchor nodes in the anchor node set can be distributed nodes whose quota threshold margin is limited by a guaranteed lower limit during quota scheduling, and whose quota threshold is prioritized for increase and last for decrease. Designated nodes in the designated node set can be distributed nodes whose quota threshold is not changed during quota scheduling, but whose quota threshold is still included in the verification of the global quota upper limit. For example, a designated node could be a server host of a market-making system that must continuously provide two-sided quotes and whose sending rate cannot be reduced. In practice, the aforementioned executing entity can, for each node role identifier in the aforementioned node role identifier set, determine the distributed node corresponding to the aforementioned node role identifier as a normal node in response to determining that the aforementioned node role identifier represents a normal node, and determine the distributed node corresponding to the aforementioned node role identifier as an anchor node in response to determining that the aforementioned node role identifier represents a specified node in response to determining that the aforementioned node role identifier represents a specified node, thereby obtaining a set of normal nodes, a set of anchor nodes, and a set of specified nodes.

[0026] The second step involves performing the following partitioning steps on the node operation information of each node in the above node operation information set to obtain the initial quota information set: Sub-step 1: Determine the quota threshold, actual occupancy, and minimum guaranteed value included in the above node operation information as the target quota threshold, target actual occupancy, and target minimum guaranteed value.

[0027] Sub-step 2: In response to determining that the distributed node corresponding to the above node operation information is a normal node, the above target quota threshold is divided according to the above target measured occupancy to obtain the node elastic quota corresponding to the above normal node. In practice, the above executing entity can determine the node elastic quota corresponding to the above normal node as the difference between the above target quota threshold and the above target measured occupancy.

[0028] Sub-step 3: In response to determining the distributed node corresponding to the above node operation information as the anchor node, the target quota threshold is divided according to the above measured target occupancy and the above guaranteed minimum value to obtain the node elastic quota corresponding to the anchor node. In practice, the executing entity can determine the node elastic quota corresponding to the anchor node as the difference between the above target quota threshold and the higher of the above measured target occupancy and the above guaranteed minimum value.

[0029] Sub-step 4: In response to determining that the distributed node corresponding to the above node operation information is a designated node, the preset fixed quota is determined as the node elastic quota corresponding to the designated node. The preset fixed quota can be a pre-set, relinquished quota threshold margin for the designated node. The preset fixed quota can be 0.

[0030] Sub-step 5: Determine the above-mentioned node elastic quota and node operation information as the initial quota information corresponding to the distributed node.

[0031] Step 103: In response to determining that the initial quota information set meets the preset over-limit condition relative to the global quota upper limit, the initial quota information set is subjected to over-limit convergence processing according to the global quota upper limit to obtain the quota information set within the constraints.

[0032] In some embodiments, the executing entity may, in response to determining that the initial quota information set satisfies a preset over-limit condition relative to the global quota upper limit, perform over-limit convergence processing on the initial quota information set according to the global quota upper limit to obtain a constraint-bound quota information set. The preset over-limit condition may be that the sum of all quota thresholds included in the initial quota information set is greater than the global quota upper limit, or it may be that there is initial quota information in the initial quota information set whose measured occupancy is greater than a quota threshold. The constraint-bound quota information in the constraint-bound quota information set may be information reflecting the quota thresholds and available margins of distributed nodes under the premise that the sum of all quota thresholds does not exceed the global quota upper limit. Optionally, in response to determining that the initial quota information set does not satisfy the preset over-limit condition, the executing entity may determine the initial quota information set as a constraint-bound quota information set.

[0033] In some optional implementations of certain embodiments, the above-mentioned process of performing over-limit convergence processing on the initial quota information set based on the global quota limit to obtain the quota information set within the constraints may include the following steps: The first step is to perform round-matching on the initial quota information set to obtain the aligned quota information set. The preset quota level can be the interval between adjacent values ​​that the quota threshold can take, and the quota threshold is an integer multiple of the preset quota level. For example, the preset quota level can be a standard flow rate quota of the aforementioned dispatch gateway. In practice, the executing entity can first determine the initial quota information in the initial quota information set whose measured usage is greater than the quota threshold and whose corresponding distributed node is not a specified node as the quota information set to be aligned. Then, for each quota information in the quota information set to be aligned, the lowest value among the integer multiples of the preset quota level that is not less than the measured usage included in the quota information to be aligned is determined as the aligned quota threshold. Finally, the quota threshold included in the quota information to be aligned is replaced with the aligned threshold. Subsequently, in response to determining that the distributed node corresponding to the quota information to be aligned is a normal node, the difference between the aligned quota threshold and the measured occupancy included in the quota information to be aligned is determined as the node elastic quota included in the quota information to be aligned. In response to determining that the distributed node corresponding to the quota information to be aligned is an anchor node, the difference between the aligned quota threshold and the higher of the measured occupancy and the minimum guaranteed value included in the quota information to be aligned is determined as the node elastic quota included in the quota information to be aligned. Finally, the replaced initial quota information set is determined as the aligned quota information set.

[0034] The second step involves filtering and sorting the distributed node set corresponding to the aligned quota information set to obtain a sequence of adjustment points to be lowered. This sequence can be composed of distributed nodes arranged in the order in which their quota thresholds are lowered. In practice, the executing entity can first identify the distributed nodes in the aligned quota information set that are not designated nodes and whose elastic quota is not less than the preset quota level as the set of adjustment points to be lowered. Then, the ordinary nodes in this set are arranged in descending order of their elastic quota to obtain a sequence of ordinary nodes. Finally, the anchor nodes in this set are arranged in descending order of their elastic quota after the sequence of ordinary nodes to obtain the final sequence of adjustment points.

[0035] The third step involves performing the following reduction steps based on the above sequence of adjustment points and the aligned quota information set: Sub-step 1: Determine the first adjustment point in the above sequence of adjustment points as the target adjustment point, and remove the target adjustment point from the above sequence of adjustment points to obtain the remaining sequence of adjustment points.

[0036] Sub-step 2 involves lowering the quota thresholds included in the aligned quota information corresponding to the target adjustment point to obtain the adjustment point quota information. It also involves replacing the aligned quota information set with the aligned quota information corresponding to the target adjustment point to obtain the adjusted quota information set. In practice, the executing entity can first determine the portion of the sum of all quota thresholds in the aligned quota information set that exceeds the global quota upper limit as the required release quota. Then, it determines the highest value among the integer multiples of the preset quota tiers that is not greater than the node elastic quota included in the aligned quota information corresponding to the target adjustment point as the releaseable quota. Next, it determines the lower of the required release quota and the releaseable quota as the target release quota. Finally, it determines the portion of the quota thresholds included in the aligned quota information corresponding to the target adjustment point that exceeds the target release quota as the adjusted quota threshold. Subsequently, in response to determining that the aforementioned target adjustment point is a normal node, the difference between the aforementioned reduced quota threshold and the measured occupancy corresponding to the aforementioned target adjustment point is determined as the node elastic quota corresponding to the aforementioned target adjustment point. In response to determining that the aforementioned target adjustment point is an anchor node, the difference between the aforementioned reduced quota threshold and the higher of the aforementioned measured occupancy and the aforementioned minimum guaranteed limit corresponding to the aforementioned target adjustment point is determined as the node elastic quota corresponding to the aforementioned target adjustment point. Finally, the quota threshold included in the aligned quota information corresponding to the aforementioned target adjustment point is replaced with the aforementioned reduced quota threshold, and the node elastic quota is replaced with the node elastic quota corresponding to the aforementioned target adjustment point, to obtain the adjustment point quota information.

[0037] Sub-step 3: In response to determining that the sum of all quota thresholds included in the reduced quota information set is less than or equal to the global quota limit, the reduced quota information set is determined as the within-constraint quota information set.

[0038] Sub-step 4: In response to determining that the sum of all quota thresholds included in the reduced quota information set is greater than the global quota limit, the reduced quota information set is determined as the aligned quota information set, and the remaining sequence of adjustment points is determined as the sequence of adjustment points to be adjusted, so as to execute the reduction step again.

[0039] Optionally, in response to determining that the remaining sequence of points to be adjusted is empty and that the sum of the various quota thresholds included in the adjusted quota information set is greater than the global quota limit, the executing entity may end the quota scheduling and send the over-limit prompt information to the operation and maintenance verification terminal.

[0040] Step 104: Perform occupancy trend identification processing on the quota information set within the constraints to obtain the supply node set.

[0041] In some embodiments, the aforementioned executing entity may perform occupancy trend identification processing on the aforementioned quota information set within constraints to obtain a supply node set. The supply nodes in the supply node set may be distributed nodes whose quota thresholds can be modified in this quota scheduling and which are not excluded from the quota relinquishment threshold.

[0042] In addressing the technical problems mentioned above, when employing technical solutions to solve these issues, the following technical problems often arise in the application scenario: differentiated quota scheduling for a distributed node set with rapidly changing transmission volumes during periods of market volatility. This is because, during periods of market volatility, the transmission volume of some distributed nodes is at the beginning of an upward trend. These nodes still have a relatively large quota threshold margin at the time of quota scheduling. When determining the remaining quota threshold margin to be ceded based on the current period's remaining volume, these nodes are identified as supply nodes, and their quota thresholds are subsequently lowered. However, in subsequent periods, the lowered quota threshold is reached, and transmission requests are delayed. A conventional solution is to set a uniform threshold for the cumulative change in transmission volume, excluding distributed nodes whose cumulative change exceeds this threshold from the supply node category. However, a single threshold is difficult to accommodate distributed nodes with significantly different quota threshold sizes, and the cumulative change must continue to rise for a considerable period before exceeding the threshold, which may not occur at the time of quota scheduling. Meanwhile, a single threshold cannot distinguish between distributed nodes that are still climbing and those whose changes have stabilized. Distributed nodes still climbing are retained as supply nodes, while those that have stabilized are excluded. This reduction in supply nodes causes a persistent mismatch between the quota threshold and the actual transmission volume of each node, leading to increased system resource consumption, higher node load, and reduced operational stability. Considering the following requirements for this application scenario: significant differences in quota thresholds among distributed nodes, small changes in transmission volume at the starting point of the increase, and similar transmission volume change patterns between continuously climbing and stabilized states, we have decided to adopt the following solution: Optionally, the above-mentioned occupancy trend identification processing of the quota information set within the constraints to obtain the supply node set may include the following steps: The first step is to perform the following trend identification steps for each constraint quota information in the above constraint quota information set: Sub-step 1 involves performing differential processing on the occupancy sequence included in the quota information within the aforementioned constraints to obtain an occupancy increment sequence. The occupancy increment in this sequence characterizes the change in the measured occupancy of a distributed node between two adjacent sampling positions. In practice, the executing entity can determine the occupancy increment as the difference between the measured occupancy at each sampling position (excluding the first position) and the measured occupancy at the previous sampling position, thus obtaining the occupancy increment sequence. The occupancy increment is negative when the measured occupancy at the sampling position is lower than the measured occupancy at the previous sampling position.

[0043] Sub-step 2 involves performing tolerance adjustment on the distributed nodes corresponding to the aforementioned quota information within the constraints, based on the aforementioned occupancy increment sequence and the quota threshold included in the aforementioned quota information within the constraints, to obtain an adaptive accumulation threshold. This adaptive accumulation threshold can be a critical value used to determine when the occupancy increment of a distributed node exhibits continuous unidirectional accumulation. In practice, the executing entity can first determine the lower quartile of the absolute value of each occupancy increment in the aforementioned occupancy increment sequence as the baseline fluctuation amplitude. This baseline fluctuation amplitude characterizes the fluctuation amplitude of the measured occupancy of the distributed node at sampling positions where continuous unidirectional accumulation has not occurred. Then, the higher of the product of the aforementioned quota threshold and the preset tolerance ratio and the aforementioned baseline fluctuation amplitude is determined as the initial drift tolerance. Next, in response to determining that the initial drift tolerance is less than the product of the quota threshold and the preset threshold ratio, the initial drift tolerance is determined as the drift tolerance of the distributed node corresponding to the quota information within the constraints. And in response to determining that the initial drift tolerance is not less than the product of the quota threshold and the preset threshold ratio, the product of the quota threshold and the preset tolerance ratio is determined as the drift tolerance of the distributed node corresponding to the quota information within the constraints. The drift tolerance can be the amount by which the incremental occupancy is allowed to be deducted item by item without being judged as continuously accumulating in the same direction. For example, the preset tolerance ratio can be 2%. Finally, the product of the quota threshold and the preset threshold ratio is determined as the adaptive accumulation threshold. For example, the preset threshold ratio can be 10%. As an example, the adaptive accumulation threshold obtained by the distributed node with the higher quota threshold is higher than the adaptive accumulation threshold obtained by the distributed node with the lower quota threshold.

[0044] Sub-step 3 involves performing a step-by-step cumulative boundary detection on the aforementioned occupancy increment sequence based on the aforementioned adaptive cumulative threshold, to obtain first change information and change point positions. The first change information characterizes whether the accumulated amount of the aforementioned occupancy increment sequence exceeds the aforementioned adaptive cumulative threshold. The first change information may include the maximum accumulated amount in the aforementioned occupancy increment sequence and the boundary determination result. The boundary determination result may be that the maximum accumulated amount exceeds the aforementioned adaptive cumulative threshold, or it may be that the maximum accumulated amount does not exceed the aforementioned adaptive cumulative threshold. In practice, the executing entity may first use an accumulation and detection method to add the accumulated amount of the previous term to the aforementioned occupancy increment for each occupancy increment in the aforementioned occupancy increment sequence to obtain an initial accumulated amount. Then, the difference between the initial accumulated amount and the aforementioned drift tolerance is determined as the deducted accumulated amount. Next, in response to determining that the deducted accumulated amount is greater than 0, the deducted accumulated amount is determined as the accumulated amount of the aforementioned occupancy increment; and in response to determining that the deducted accumulated amount is less than or equal to 0, 0 is determined as the accumulated amount of the aforementioned occupancy increment. The aforementioned accumulation and detection methods may include, but are not limited to, at least one of the following: the CUSUM algorithm and the Page-Hinkley algorithm. Then, the position of the first occupancy increment exceeding the aforementioned adaptive accumulation threshold is determined as the change point position. If there are no occupancy increments in the aforementioned occupancy increment sequence that exceed the aforementioned adaptive accumulation threshold, the last position of the aforementioned occupancy increment sequence is determined as the change point position. Finally, the maximum occupancy among the obtained occupancy values, and the comparison result between the maximum occupancy and the aforementioned adaptive accumulation threshold, are determined as the first change information.

[0045] Sub-step 4: Based on the aforementioned change point positions, segment the aforementioned occupancy increment sequence to obtain the rising segment increment sequence. In practice, the aforementioned execution entity can determine each occupancy increment from the aforementioned change point position to the last position in the aforementioned occupancy increment sequence as the rising segment increment sequence.

[0046] Sub-step 5 involves detecting the rate of change of the aforementioned rising segment increment sequence to obtain second change information. This second change information characterizes the rate of change of the aforementioned rising segment increment sequence relative to a preset rate threshold. The second change information may include the aforementioned change rate and a rate determination result. The change rate characterizes the average change in the measured occupancy of the distributed node at each sampling position from the aforementioned change point. The rate determination result may indicate that the change rate exceeds the preset rate threshold, or that the change rate does not exceed the preset rate threshold. The preset rate threshold may be the ratio of the adaptive cumulative threshold to the number of occupancy increments included in the occupancy increment sequence. In practice, the executing entity may first determine the average value of each occupancy increment in the aforementioned rising segment increment sequence as the change rate. Then, the aforementioned change rate, and the comparison result between the aforementioned change rate and the preset rate threshold, are determined as the second change information.

[0047] Sub-step 6: Based on the number of items between the aforementioned change point position and the last position of the aforementioned occupancy increment sequence, perform a rising timeliness measurement on the distributed nodes corresponding to the aforementioned quota information within the constraints to obtain a rising timeliness value. The rising timeliness value characterizes the number of sampling positions traversed by the distributed node's occupancy from its initial continuous accumulation in the same direction to the current sampling position. In practice, the executing entity can determine the rising timeliness value as the number of items between the aforementioned change point position and the last position of the aforementioned occupancy increment sequence. As an example, when the aforementioned occupancy increment sequence includes 11 occupancy increments and the aforementioned change point position is the 7th item, the aforementioned rising timeliness value is 4.

[0048] The second step involves performing trend-based diversion processing on the distributed node set corresponding to the aforementioned quota information set within constraints, based on the obtained first change information set, second change information set, and rising timeliness value set, to obtain a supply node set. In practice, the executing entity can first identify the distributed nodes in the distributed node set corresponding to the aforementioned quota information set within constraints whose cross-boundary judgment result in the corresponding first change information is that the maximum cumulative amount exceeds the aforementioned adaptive cumulative threshold, as rising nodes. Then, among the rising nodes, those whose rate judgment result in the corresponding second change information is that the change rate exceeds the aforementioned preset rate threshold, and those whose rising timeliness value is less than or equal to the preset timeliness threshold, are identified as exclusion nodes. The aforementioned preset timeliness threshold can be a pre-set critical value for determining whether the increase in the occupancy of a distributed node is in the initial stage. For example, the aforementioned preset timeliness threshold can be 3. Finally, the distributed nodes in the distributed node set corresponding to the aforementioned quota information set within constraints, excluding the aforementioned exclusion nodes and designated nodes, are identified as supply nodes, thus obtaining a supply node set.

[0049] The above technical content, combined with step "Step 107" below, serves as an inventive point of this disclosure, solving the technical problem of "increased system resource consumption, increased node load, and reduced operational stability of distributed nodes." Factors leading to increased system resource consumption, increased node load, and reduced operational stability of distributed nodes are often as follows: Distributed nodes whose transmission volume is at the starting point of an increase still have a large quota threshold margin at the quota scheduling time. When determining the remaining quota threshold margin that can be ceded based on the remaining amount in the current period, they are identified as supply nodes, and the quota threshold is subsequently lowered. However, in subsequent periods, the lowered quota threshold is reached, and the request to send is delayed. Furthermore, when setting a uniform threshold for the cumulative change in transmission volume, the same threshold is difficult to adapt to distributed nodes with large differences in quota threshold size. The cumulative change has not yet exceeded the threshold at the quota scheduling time, and it is impossible to distinguish between distributed nodes that are still climbing and those whose changes have stabilized. The number of supply nodes decreases due to excessive exclusion, resulting in a continuous mismatch between the quota threshold and the actual transmission volume of each node. Solving these factors can reduce the system resource consumption of distributed nodes, lower node load, and improve operational stability. To achieve this effect, this disclosure first performs differential processing on the occupancy sequence included in the quota information within the constraints to obtain an occupancy increment sequence. Here, the occupancy increment sequence reflects the change in transmission volume of distributed nodes between two adjacent sampling positions, providing a data basis for detecting the continuous unidirectional accumulation of transmission volume. Next, based on the occupancy increment sequence and the quota threshold included in the quota information within the constraints, tolerance adjustment processing is performed on the distributed nodes to obtain an adaptive accumulation threshold. Here, the adaptive accumulation threshold and drift tolerance are determined according to the quota threshold of each distributed node, so that distributed nodes with different quota threshold sizes are subject to a judgment scale commensurate with their quota size. Then, based on the adaptive accumulation threshold, the occupancy increment sequence is accumulated item by item for boundary detection to obtain the first change information and the change point position. Here, the occupancy increment after drift tolerance deduction is accumulated item by item, so that normal fluctuations are offset item by item, and small increments in the continuous unidirectional direction are accumulated, and the increase in transmission volume is detected at the sampling position where the accumulated volume first exceeds the boundary. Finally, the occupancy increment sequence is segmented according to the change point position to obtain the rising segment increment sequence. The rate of change of the rising segment increment sequence is detected to obtain the second change information. The rising timeliness of the distributed nodes is measured according to the number of terms between the change point position and the last position of the occupancy increment sequence to obtain the rising timeliness value. The distributed node set is then subjected to trend diversion processing based on the obtained first change information set, the obtained second change information set, and the obtained rising timeliness value set to obtain the supply node set.Here, the rate of change and the rise duration respectively characterize the speed of the rise and the number of sampling positions where the rise has been ongoing. This ensures that distributed nodes with a still high rate of change or those in the initial stage of the rise are excluded from the supply nodes, while distributed nodes whose rise has been ongoing for a long time and whose rate of change has declined are retained as supply nodes, preventing an excessive reduction in the number of supply nodes due to exclusion. The above technical content, combined with step "Step 107" below, can reduce the system resource consumption of distributed nodes, lower node load, and improve operational stability.

[0050] Step 105: Based on the supply node set, perform steady-state redistribution processing on the constrained quota information set to obtain the target quota information set.

[0051] In some embodiments, the executing entity may perform steady-state redistribution processing on the constrained quota information set based on the supply node set to obtain a target quota information set. The target quota information in the target quota information set may reflect the quota thresholds to be set for distributed nodes and the available margins they can relinquish.

[0052] In some optional implementations of certain embodiments, the above-mentioned steady-state redistribution processing of the constrained quota information set based on the above-mentioned supply node set to obtain the target quota information set may include the following steps: The first step involves classifying the distributed node set corresponding to the aforementioned quota information set within the constraints according to a preset buffer limit, resulting in a set of adjustment points. The preset buffer limit can be the quota threshold remaining amount that a distributed node should retain after this quota scheduling. The adjustment points in the set of adjustment points can be distributed nodes whose elastic quota is less than the preset buffer limit. In practice, the executing entity can first determine the preset buffer limit for each quota information within the constraints by multiplying the quota threshold included in the quota information by a preset buffer ratio. For example, the preset buffer ratio could be 15%. Then, distributed nodes in the set of distributed nodes corresponding to the aforementioned quota information set within the constraints that are not specified nodes and whose elastic quota is less than the preset buffer limit are identified as adjustment points, thus obtaining the set of adjustment points.

[0053] The second step involves performing quota inter-adjustment processing on the constrained quota information set based on the aforementioned set of demand adjustment points and the aforementioned set of supply nodes to obtain the target quota information set. In practice, the executing entity can first determine the gap difference value corresponding to the demand adjustment point as the difference between the preset buffer quota and the node elastic quota corresponding to each demand adjustment point in the aforementioned set of demand adjustment points, and determine the lowest value among integer multiples of the aforementioned preset quota levels that is not less than the gap difference value as the node gap quota corresponding to the aforementioned demand adjustment point. Then, for each supply node in the aforementioned set of supply nodes, determine the surplus difference value corresponding to the aforementioned supply node as the difference between the node elastic quota and the preset buffer quota corresponding to the aforementioned supply node, and determine the highest value among integer multiples of the aforementioned preset quota levels that is not greater than the surplus difference value as the inter-adjustable quota corresponding to the aforementioned supply node. Next, for each adjustment point in the aforementioned adjustment point set, the lower of the adjustable quota corresponding to the supply node with an adjustable quota greater than 0 in the aforementioned supply node set and the node gap quota corresponding to the adjustment point is determined as the current adjustment quota. The quota threshold corresponding to the supply node is reduced by the current adjustment quota, and the quota threshold corresponding to the adjustment point is increased by the current adjustment quota, resulting in the adjusted quota threshold. Then, for each distributed node whose quota threshold has been reduced or increased, in response to determining the distributed node as a normal node, the difference between the adjusted quota threshold and the measured occupancy of the distributed node is determined as the node elasticity quota for the distributed node. In response to determining the distributed node as an anchor node, the difference between the adjusted quota threshold and the higher of the measured occupancy and the minimum guaranteed limit is determined as the node elasticity quota for the distributed node. Finally, the aforementioned constrained quota information set is determined as the target quota information set.

[0054] In the process of adopting technical solutions to address the technical problems mentioned above, the following technical issues often arise in the application scenario: during periods of market volatility, the peak sending times of some distributed node combinations overlap while the peak sending times of other node combinations are staggered (for example, in multiple order placement systems carrying different securities trading businesses, the peak order placement times of some order placement systems fall at the same time, while the peak order placement times of other order placement systems fall at different times). Since the quota threshold margin reflects the sending volume of distributed nodes at the time of quota scheduling, it does not reflect the peak sending times in subsequent periods. When adjusting quota thresholds based on the level of the quota threshold margin, it is not determined whether the node giving up the margin and the node receiving the margin belong to a node combination with staggered sending peaks. The two may belong to a node combination with overlapping sending peaks, and both may surge in the same period in the subsequent period. Both may be insufficient in the period, and the adjusted quota threshold will be reversed in the next scheduling cycle. The conventional solution is to determine whether the receiving node is at a low position during the peak sending period of the node relinquishing its transmission capacity, and then decide whether to inter-swap based on this. However, this does not consider whether the nodes relinquishing their transmission capacity during the peak sending period of the receiving node are simultaneously surging, and the duration of the peak sending periods of the two distributed nodes differs, resulting in varying degrees of overlap in the two directions, with reverse overlap being missed. Furthermore, there are few node combinations with completely staggered sending peaks. If all nodes are discarded based solely on whether their peaks are staggered, the number of inter-swapping node combinations is too small, and the quota threshold margin cannot be adequately adjusted, causing a continuous mismatch between the quota threshold and the actual sending volume of each node. This leads to increased system resource consumption, higher node load, and reduced operational stability for the distributed nodes. Considering the following requirements of this application scenario: varying degrees of overlap in the sending peaks of different node combinations; varying degrees of overlap between the sending peaks of the two distributed nodes; and a limited number of node combinations with completely staggered sending peaks, we have decided to adopt the following solution: Optionally, the above-mentioned steady-state redistribution of the constrained quota information set based on the above-mentioned supply node set to obtain the target quota information set may include the following steps: The first step involves identifying buffer gaps in the aforementioned quota information set within constraints, based on a preset buffer limit, to obtain a set of adjustment points and a set of node gap limits. The node gap limit in the node gap limit set can be the quota threshold that needs to be increased at each adjustment point to ensure that the quota threshold margin is not lower than the preset buffer limit. In practice, the executing entity can first determine the preset buffer limit for the corresponding distributed node for each quota information within constraints in the aforementioned quota information set by multiplying the quota threshold included in the quota information by a preset buffer ratio. For example, the preset buffer ratio could be 15%. Then, distributed nodes in the distributed node set corresponding to the aforementioned quota information set that are not specified nodes and whose node elasticity limit is less than the preset buffer limit are identified as adjustment points, thus obtaining a set of adjustment points. Finally, for each adjustment point in the above set of adjustment points, the difference between the preset buffer quota and the node elastic quota corresponding to the adjustment point is determined as the gap difference value corresponding to the adjustment point, and the lowest value among the integer multiples of the preset quota level that is not less than the gap difference value is determined as the node gap quota corresponding to the adjustment point, thus obtaining the node gap quota set.

[0055] The second step involves extracting occupancy segments from the aforementioned supply node set and demand adjustment point set, resulting in a supply occupancy segment information set and a demand adjustment point occupancy segment information set. The occupancy rate can be the proportion of the measured occupancy amount within the quota threshold. An occupancy segment can be a section in the occupancy sequence consisting of one or more consecutive sampling positions where the occupancy rate is greater than a preset high occupancy ratio. The supply occupancy segment information in the supply occupancy segment information set and the demand adjustment point occupancy segment information in the demand adjustment point occupancy segment information set can be composed of the start and end sampling positions of each occupancy segment corresponding to the distributed node. In practice, the executing entity can first, for each distributed node in the aforementioned supply node set and demand adjustment point set, determine the occupancy rate at each sampling position as the ratio of the measured occupancy amount at each sampling position in the occupancy sequence included in the constraint quota information corresponding to the distributed node to the quota threshold included in the constraint quota information. Then, sampling positions with occupancy rates greater than the preset high occupancy ratio are designated as high occupancy positions. For example, the preset high occupancy rate can be 70%. Finally, the first and last high occupancy positions in each consecutive high occupancy position are determined as the start and end sampling positions of the occupancy segment, respectively, to obtain the supply occupancy segment information or the occupancy segment information of the adjustment point corresponding to the above distributed node.

[0056] The third step involves performing bidirectional peak-shifting matching on the aforementioned supply occupancy segment information set and the aforementioned demand adjustment point occupancy segment information set to obtain a peak-shifting matching degree set and a high occupancy overlap rate set. The peak-shifting matching degree in the peak-shifting matching degree set characterizes the degree to which the sampling positions covered by the occupancy segments of one of the supply nodes and the demand adjustment points are not covered by the occupancy segments of the other. The high occupancy overlap rate in the high occupancy overlap rate set characterizes the proportion of sampling positions simultaneously covered by the occupancy segments of both the supply node and the demand adjustment point in all sampling positions. In practice, the executing entity can first, for each node pair consisting of each supply node in the supply node set and each demand adjustment point in the demand adjustment point set, determine the first staggering ratio of the node pair as the ratio of the number of sampling positions not covered by the occupancy segments of the demand adjustment point among the sampling positions covered by each occupancy segment of the supply node to the number of sampling positions covered by each occupancy segment of the supply node. Then, the ratio of the number of sampling positions not covered by the occupied segments of the aforementioned adjustment points to the number of sampling positions covered by the occupied segments of the aforementioned adjustment points is determined as the second offset ratio of the node pair. When the aforementioned supply node has no occupied segments, the aforementioned first offset ratio is determined to be 1; when the aforementioned adjustment point has no occupied segments, the aforementioned second offset ratio is determined to be 1. Next, the lower of the aforementioned first offset ratio and the aforementioned second offset ratio is determined as the peak matching degree of the aforementioned node pair, thus obtaining the peak matching degree set. Finally, the ratio of the number of sampling positions simultaneously covered by the occupied segments of the aforementioned supply node and the aforementioned adjustment point to the aforementioned preset sampling number is determined as the high occupancy overlap rate of the aforementioned node pair, thus obtaining the high occupancy overlap rate set.

[0057] The fourth step involves performing node association filtering on the aforementioned peak-shifting matching degree set and the aforementioned high occupancy overlap rate set, thereby obtaining a node association information set. The node association information in this set can represent the quota threshold margin that supply nodes can relinquish to demand-side adjustment points. In practice, the executing entity can identify node pairs with a peak-shifting matching degree not less than a preset matching degree threshold and a high occupancy overlap rate not greater than a second preset overlap ratio as node association information, thus obtaining the node association information set. For example, the preset matching degree threshold could be 60%, and the second preset overlap ratio could be 20%.

[0058] Fifth, based on the aforementioned node association information set, node gap quota set, and constrained quota information set, quota inter-adjustment mapping is performed on the aforementioned supply node set and demand adjustment point set to obtain an initial quota inter-adjustment matrix. This initial quota inter-adjustment matrix can be a two-dimensional array obtained by arranging the quota threshold margins relinquished by each supply node to each demand adjustment point according to the correspondence between supply nodes and demand adjustment points. In practice, the executing entity can first, for each supply node in the aforementioned supply node set, determine the difference between the node elasticity quota corresponding to the supply node and the preset buffer quota corresponding to the supply node as the margin difference value corresponding to the supply node, and determine the highest value among integer multiples of the preset quota level that is not greater than the margin difference value as the inter-adjustable quota corresponding to the supply node. Then, the node association information in the aforementioned node association information set is arranged in the order that the corresponding demand adjustment point is an anchor node first, followed by a regular node, and the corresponding supply node is a regular node first, followed by an anchor node, to obtain a node association information sequence. Next, for each node association information in the above node association information sequence, the lower of the unallocated portion of the inter-adjustable quota of the supply node corresponding to the above node association information and the unfilled portion of the node gap quota of the node requiring adjustment corresponding to the above node association information is determined as the initial inter-adjustment quota corresponding to the above node association information. Finally, each of the above initial inter-adjustment quotas is mapped to the corresponding position in the two-dimensional array according to the above correspondence between the supply node and the above requiring adjustment point, and the initial inter-adjustment quota at the corresponding position where there is no node association information is determined as 0, to obtain the initial quota inter-adjustment matrix.

[0059] Step 6: Based on the aforementioned high occupancy overlap rate set, the initial quota intermodulation matrix is ​​subjected to tiered limiting processing to obtain the target quota intermodulation matrix. The target quota intermodulation matrix can be a two-dimensional array obtained by arranging the initial intermodulation quotas according to the correspondence between supply nodes and demand adjustment points after tiering. In practice, the executing entity can first, for each initial intermodulation quota in the initial quota intermodulation matrix, in response to determining that the high occupancy overlap rate of the node pair corresponding to the initial intermodulation quota is not greater than a first preset overlap ratio, determine the initial intermodulation quota as the target intermodulation quota. Then, in response to determining that the high occupancy overlap rate is greater than the first preset overlap ratio, determine the lower of the product of the preset tier limit number and the preset quota tier, and the initial intermodulation quota, as the target intermodulation quota. For example, the first preset overlap ratio can be 5%, and the preset tier limit number can be 1. Finally, map each of the target intermodulation quotas to the corresponding positions in the two-dimensional array according to the correspondence between supply nodes and demand adjustment points to obtain the target quota intermodulation matrix. When the target inter-transfer quota is less than the corresponding initial inter-transfer quota, the actual inter-transfer quota threshold between the corresponding node combinations is less than the amount determined by the node gap quota.

[0060] Step 7: Based on the aforementioned target quota inter-adjustment matrix, update the quota information set within the constraints to obtain the target quota information set. In practice, the executing entity can first, for each target inter-adjustment amount in the target quota inter-adjustment matrix, reduce the quota threshold of the supply node corresponding to the target inter-adjustment amount by the target inter-adjustment amount, and increase the quota threshold of the adjustment point corresponding to the target inter-adjustment amount by the target inter-adjustment amount, to obtain the inter-adjusted quota threshold. Then, for each distributed node whose quota threshold has been reduced or increased, in response to determining that the distributed node is a normal node, the difference between the inter-adjusted quota threshold corresponding to the distributed node and the measured occupancy corresponding to the distributed node is determined as the node elasticity quota corresponding to the distributed node; and in response to determining that the distributed node is an anchor node, the difference between the inter-adjusted quota threshold corresponding to the distributed node and the higher of the measured occupancy and the minimum guaranteed limit corresponding to the distributed node is determined as the node elasticity quota corresponding to the distributed node. Finally, the above-mentioned set of quota information within constraints is determined as the target quota information set.

[0061] The above technical content, combined with step "107" below, serves as an inventive point of this disclosure, solving the technical problem of "increased system resource consumption, increased node load, and reduced operational stability of distributed nodes". Factors leading to increased system resource consumption, higher node load, and reduced operational stability in distributed nodes are often as follows: Quota threshold margins reflect the transmission volume of distributed nodes at the time of quota scheduling, but do not reflect their transmission peaks in subsequent periods. When adjusting quota thresholds based on margin levels, it is not determined whether the nodes relinquishing and receiving margins belong to a combination with staggered transmission peaks. They may belong to a combination with overlapping transmission peaks, both surging in the same subsequent period, resulting in insufficient capacity for both. The adjusted quota threshold is then reversed in the next scheduling cycle. Furthermore, if only the transmission peak of the node relinquishing the margin is checked to see if the receiving node is at a low level, such overlaps are missed. And if all nodes are discarded based solely on whether their transmission peaks are staggered, the number of interchangeable node combinations is too small, and the quota threshold margin is not adequately adjusted, causing a continuous mismatch between the quota threshold and the actual transmission volume of each node. Solving these factors can reduce system resource consumption, lower node load, and improve operational stability in distributed nodes. To achieve this effect, this disclosure first performs buffer gap identification processing on the quota information set within constraints based on a preset buffer limit, obtaining a set of adjustment points to be adjusted and a set of node gap limits. Then, based on the set of supply nodes and the set of adjustment points to be adjusted, it performs occupancy segment extraction processing on the quota information set within constraints, obtaining a set of supply occupancy segments and a set of adjustment point occupancy segments. Here, the occupancy segment information, with its start and end sampling positions, indicates the location of the peak transmission times of each distributed node within the scheduling period, providing a data basis for comparing the distribution of peak transmission times among distributed nodes. Next, the supply occupancy segment information set and the adjustment point occupancy segment information set undergo bidirectional peak-shifting matching processing, obtaining a peak-shifting matching degree set and a high occupancy overlap rate set. Here, using the sampling positions covered by the occupancy segments of supply nodes as a benchmark and the sampling positions covered by the occupancy segments of adjustment points as a benchmark, the offset ratios in two directions are obtained respectively, and the lower of the two is taken, so that node combinations offset in only one direction are not judged as mutually offset. Then, based on the peak matching degree set and the high occupancy overlap rate set, node association screening is performed on the supply node set and the demand adjustment point set to obtain the node association information set. Furthermore, based on the node association information set, the node gap quota set, and the constraint quota information set, quota inter-adjustment mapping is performed on the supply node set and the demand adjustment point set to obtain the initial quota inter-adjustment matrix. Here, node combinations with staggered transmission peaks are screened out pairwise, and the quota threshold margin is adjusted among distributed nodes whose transmission peaks do not overlap, ensuring that the adjusted quota threshold is actually available to nodes with receiving margins in subsequent time periods.Finally, based on the high overlap rate set, the initial quota intermodulation matrix is ​​subjected to tiered limiting processing to obtain the target quota intermodulation matrix. Then, based on the target quota intermodulation matrix, the quota information set within the constraints is updated to obtain the target quota information set. Here, node combinations with partially overlapping transmission peaks are not excluded outright. Instead, the quota threshold for intermodulation is limited to a preset tiered limit based on the degree of overlap. This ensures that the number of intermodulation-capable node combinations is not too small due to outright rejection, and that when partially overlapping node combinations simultaneously surge in subsequent periods, the added quota threshold is insufficient to cause both to significantly exceed their respective quota thresholds. The above technical content, combined with step "Step 107" below, can reduce the system resource consumption of distributed nodes, lower node load, and improve operational stability.

[0062] Step 106: Based on the target quota information set and the initial quota information set, perform change determination processing on the distributed node set to obtain the quota change information set.

[0063] In some embodiments, the executing entity may perform change determination processing on the distributed node set based on the target quota information set and the initial quota information set to obtain a quota change information set. The quota change information in the quota change information set may reflect changes in the quota thresholds of the distributed nodes. The quota change information may include a change type. The change type may be unchanged, decreasing, or increasing.

[0064] In some optional implementations of certain embodiments, the process of determining the change of the distributed node set based on the target quota information set and the initial quota information set to obtain the quota change information set may include the following steps: For each distributed node in the above distributed node set, perform the following change determination steps to obtain the quota change information set: Sub-step 1: Determine the quota threshold included in the target quota information corresponding to the above-mentioned distributed nodes as the quota threshold after scheduling, and determine the quota threshold included in the initial quota information corresponding to the above-mentioned distributed nodes as the initial quota threshold.

[0065] Sub-step 2: Based on the aforementioned post-scheduled quota threshold and the aforementioned initial quota threshold, determine the quota change information corresponding to the aforementioned distributed node. In practice, the executing entity can first, in response to determining that the post-scheduled quota threshold is equal to the aforementioned initial quota threshold, determine the unchanged type as the change type corresponding to the aforementioned distributed node; in response to determining that the post-scheduled quota threshold is less than the aforementioned initial quota threshold, determine the downward adjustment type as the change type corresponding to the aforementioned distributed node; and in response to determining that the post-scheduled quota threshold is greater than the aforementioned initial quota threshold, determine the upward adjustment type as the change type corresponding to the aforementioned distributed node. Finally, determine the aforementioned change type as the quota change information corresponding to the aforementioned distributed node.

[0066] Step 107: Perform batch load control on the distributed node set based on the quota change information set and the target quota information set.

[0067] In some embodiments, the execution entity may perform batch load control on the distributed node set based on the quota change information set and the target quota information set. The load control may include at least one of the following: token bucket rate limiting on the number of requests sent by the distributed nodes per unit time, and thread pool and connection pool concurrency limiting on the number of requests processed simultaneously by the distributed nodes.

[0068] In some optional implementations of certain embodiments, the process of performing batch load control on the distributed node set based on the quota change information set and the target quota information set may include the following steps: The first step involves batch-dividing the distributed node set based on the aforementioned quota change information set, resulting in a set of nodes for quota reduction and a set of nodes for quota increase. The nodes in the set of nodes for quota reduction can be distributed nodes whose quota thresholds have been lowered in this quota scheduling. Similarly, the nodes in the set of nodes for quota increase can be distributed nodes whose quota thresholds have been raised in this quota scheduling. In practice, the executing entity can first identify distributed nodes in the aforementioned distributed node set whose quota change information includes a reduction type as the reduction batch node set. Then, it can identify distributed nodes in the aforementioned distributed node set whose quota change information includes an increase type as the increase batch node set, resulting in a set of nodes for quota increase. Distributed nodes in the aforementioned distributed node set whose quota change information includes an unchanged type do not belong to either the reduction batch node set or the increase batch node set.

[0069] The second step is to perform load control on the set of nodes in the reduced quota batch based on the aforementioned target quota information set. In practice, the execution entity can perform the following adjustment steps for each node in the reduced quota batch set: First, call the management interface of the reduced quota batch node to adjust the token filling rate of the token bucket of the reduced quota batch node to the quota threshold included in the target quota information corresponding to the reduced quota batch node. The management interface may include, but is not limited to, at least one of the following: the MXBean interface provided by the Java management extension, or the endpoint provided by Spring Boot Actuator. Then, the product of the quota threshold and the average request processing time is used to determine the number of concurrent requests for the reduced quota batch node. The number of concurrent requests may be the upper limit of the number of requests that a distributed node needs to send requests according to the quota threshold, and that are simultaneously in the state of being sent but not yet receiving a response. The average request processing time may be the average time elapsed from sending a request to receiving a response by a distributed node, which can be collected by the aforementioned metric tracking component and written into the aforementioned time-series database. Finally, the management interface for the aforementioned batch node reduction is called to adjust the maximum number of threads in the thread pool and the maximum number of connections in the connection pool of the aforementioned batch node reduction to the aforementioned number of concurrent requests.

[0070] The third step involves, in response to the determination that the aforementioned set of nodes in the downgraded batch has completed load control, performing load control on the aforementioned set of nodes in the upgraded batch according to the aforementioned target quota information set. In practice, the executing entity can, in response to receiving a successful parameter adjustment response from each node in the aforementioned set of nodes in the downgraded batch, perform the following adjustment steps for each node in the aforementioned set of nodes in the upgraded batch: First, call the management interface of the aforementioned node in the upgraded batch to adjust the token filling rate of the token bucket of the aforementioned node in the upgraded batch to the quota threshold included in the target quota information corresponding to the aforementioned node in the upgraded batch. Then, the product of the quota threshold corresponding to the aforementioned node in the upgraded batch and the aforementioned average request processing time is used to determine the number of concurrent requests for the aforementioned node in the upgraded batch. Finally, call the management interface of the aforementioned node in the upgraded batch to adjust the maximum number of threads in the thread pool and the maximum number of connections in the connection pool of the aforementioned node in the upgraded batch to the aforementioned number of concurrent requests.

[0071] The above embodiments of this disclosure have the following beneficial effects: The distributed node control method based on differentiated quota scheduling in some embodiments of this disclosure can reduce the system resource consumption of distributed nodes, reduce node load, and improve operational stability. Specifically, the reason for the increased system resource consumption, increased node load, and decreased operational stability of distributed nodes is that the quota threshold of each node is uniformly re-determined according to a fixed proportion or the single-point transmission volume of the current period. The transmission rate of some nodes is continuously limited, resulting in request backlog, while the quota threshold of other nodes is continuously higher than the actual transmission volume, causing a continuous mismatch between the quota threshold and the actual transmission volume of each node, leading to increased system resource consumption, increased node load, and decreased operational stability of distributed nodes. Based on this, the distributed node control method based on differentiated quota scheduling in some embodiments of this disclosure can first, in response to receiving a quota scheduling request for a distributed node set, obtain the node operation information set and the global quota upper limit corresponding to the distributed node set. Here, the aforementioned node operation information set provides the quota threshold, measured occupancy, and occupancy sequence for each distributed node. The aforementioned global quota upper limit provides the constraints on the sum of each quota threshold, providing a data foundation for subsequent allocation of cedable quota threshold reserves and determination of over-limit. Next, based on the aforementioned node operation information set, the aforementioned distributed node set undergoes elastic quota allocation processing to obtain an initial quota information set. Here, cedable quota threshold reserves are allocated separately for ordinary nodes, anchor nodes, and designated nodes, ensuring that the quota occupied by nodes that must retain basic sending capacity and nodes whose quota thresholds are not modified are not taken as cedable amounts. The resulting node elastic quota can quantify the actual cedable reserves that each distributed node can give up in this quota scheduling. Secondly, in response to determining that the aforementioned initial quota information set meets the preset over-limit conditions relative to the aforementioned global quota upper limit, the aforementioned initial quota information set undergoes over-limit convergence processing based on the aforementioned global quota upper limit to obtain a constrained quota information set. Here, ordinary nodes with larger quota threshold margins are prioritized to relinquish their margins according to tiers, bringing the sum of all quota thresholds back below the aforementioned global quota upper limit, while simultaneously reducing the number of nodes whose quota thresholds are overwritten. Then, the quota information set within the constraints is processed for occupancy trend identification to obtain the supply node set. Here, distributed nodes whose occupancy is at the starting point of an increase are excluded from the supply nodes to prevent their quota thresholds from being reached in subsequent periods; simultaneously, distributed nodes whose occupancy has been increasing for a long time and whose rate of change has slowed down are retained within the supply nodes, so that the available margin for relinquishing is not excessively reduced due to exclusion. Afterwards, based on the supply node set, the quota information set within the constraints is processed for steady-state redistribution to obtain the target quota information set. Here, supply nodes retain a certain quota threshold margin after relinquishing, and the quota threshold margin requiring adjustment is replenished, so that the quota thresholds match the actual transmission volume of each node, and the amount of quota transferred in is reduced from being reversed in the next scheduling cycle.Subsequently, based on the aforementioned target quota information set and the aforementioned initial quota information set, the distributed node set undergoes change determination processing to obtain a quota change information set. Here, the quota thresholds before and after scheduling are compared node by node, ensuring that distributed nodes whose quota thresholds have not changed are not included in subsequent load control, thus reducing the number of nodes whose local rate limiting parameters and concurrency parameters are reset. Finally, based on the aforementioned quota change information set and the aforementioned target quota information set, the distributed node set is subjected to batch load control. Here, load control is first applied to distributed nodes whose quota thresholds have been lowered, and then to distributed nodes whose quota thresholds have been raised, ensuring that the sum of all quota thresholds never exceeds the aforementioned global quota upper limit throughout the entire load control process. Therefore, this distributed node control method based on differentiated quota scheduling can match the quota thresholds with the actual transmission volume of each node, thereby reducing the system resource consumption of distributed nodes, lowering node load, and improving operational stability.

[0072] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a distributed node control device based on differentiated quota scheduling. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this distributed node control device based on differentiated quota scheduling can be specifically applied to various electronic devices.

[0073] like Figure 2As shown, a distributed node control device 200 based on differentiated quota scheduling includes: an acquisition unit 201, a partitioning unit 202, a convergence unit 203, an identification unit 204, a reallocation unit 205, a determination unit 206, and a control unit 207. The acquisition unit 201 is configured to: in response to receiving a quota scheduling request for a distributed node set, acquire a set of node operation information and the global quota limit corresponding to the distributed node set; the partitioning unit 202 is configured to: perform elastic quota partitioning processing on the distributed node set according to the node operation information set to obtain an initial quota information set; the convergence unit 203 is configured to: in response to determining that the initial quota information set meets a preset over-limit condition relative to the global quota limit, perform over-limit convergence processing on the initial quota information set according to the global quota limit to obtain an in-constraint quota information set; the identification unit 204 is configured to: The allocation unit 204 is configured to: perform occupancy trend identification processing on the above-mentioned quota information set within constraints to obtain a supply node set; the redistribution unit 205 is configured to: perform steady-state redistribution processing on the above-mentioned quota information set within constraints based on the above-mentioned supply node set to obtain a target quota information set; the determination unit 206 is configured to: perform change determination processing on the above-mentioned distributed node set based on the above-mentioned target quota information set and the above-mentioned initial quota information set to obtain a quota change information set; and the control unit 207 is configured to: perform batch load control on the above-mentioned distributed node set based on the above-mentioned quota change information set and the above-mentioned target quota information set.

[0074] It is understandable that the units described in the distributed node control device 200 based on differentiated quota scheduling and the reference Figure 1 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the distributed node control device 200 based on differentiated quota scheduling and the units contained therein, and will not be repeated here.

[0075] The following is for reference. Figure 3 It shows a method suitable for implementing this disclosure. Figure 1 A schematic diagram of the structure of an electronic device 300 in some corresponding embodiments. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of this disclosure. Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory 302 or a program loaded from a storage device 308 into a random access memory 303. The random access memory 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, the read-only memory 302, and the random access memory 303 are interconnected via a bus 304. An input / output interface 305 is also connected to the bus 304.

[0076] Typically, the following devices can be connected to the input / output interface 305: input devices 306 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 307 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 308 including, for example, magnetic tape, hard disk, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0077] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 309, or installed from a storage device 308, or installed from a read-only memory 302. When the computer program is executed by the processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0078] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0079] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0080] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs. When the electronic device executes the aforementioned one or more programs, the electronic device causes the following actions: In response to receiving a quota scheduling request for a distributed node set, it obtains a set of node operation information and a global quota limit corresponding to the distributed node set; based on the set of node operation information, it performs elastic quota allocation processing on the distributed node set to obtain an initial quota information set; in response to determining that the initial quota information set meets a preset over-limit condition relative to the global quota limit, it performs over-limit convergence processing on the initial quota information set based on the global quota limit to obtain a constrained quota information set; it performs occupancy trend identification processing on the constrained quota information set to obtain a supply node set; based on the supply node set, it performs steady-state redistribution processing on the constrained quota information set to obtain a target quota information set; based on the target quota information set and the initial quota information set, it performs change determination processing on the distributed node set to obtain a quota change information set; and based on the quota change information set and the target quota information set, it performs batch load control on the distributed node set.

[0081] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0083] The units described in some embodiments of this disclosure can be implemented in software or in hardware. The described units can also be located in a processor.

[0084] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0085] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A distributed node control method based on differentiated quota scheduling, characterized in that, include: In response to receiving a quota scheduling request for a distributed node set, obtain the node running information set and the global quota limit corresponding to the distributed node set; Based on the node operation information set, the distributed node set is subjected to elastic quota allocation processing to obtain an initial quota information set; In response to determining that the initial quota information set satisfies a preset over-limit condition relative to the global quota upper limit, the initial quota information set is subjected to over-limit convergence processing according to the global quota upper limit to obtain a quota information set within constraints. The quota information set within the constraints is processed to identify occupancy trends, resulting in a supply node set. Based on the supply node set, the constrained quota information set is subjected to steady-state redistribution processing to obtain the target quota information set; Based on the target quota information set and the initial quota information set, the distributed node set is modified to obtain a quota change information set; Based on the quota change information set and the target quota information set, batch load control is performed on the distributed node set, wherein the batch load control of the distributed node set based on the quota change information set and the target quota information set includes: Based on the quota change information set, the distributed node set is divided into batches to obtain a set of nodes for downward adjustment and a set of nodes for upward adjustment. Based on the target quota information set, load control is performed on the set of nodes in the downgrade batch; In response to determining that the downgraded batch of nodes has completed load control, load control is performed on the upgraded batch of nodes based on the target quota information set.

2. The method of claim 1, wherein, The step of performing elastic quota allocation processing on the distributed node set based on the node operation information set to obtain an initial quota information set includes: Based on the node role identifier set included in the node operation information set, the distributed node set is classified to obtain ordinary node set, anchored node set and designated node set; The following partitioning steps are performed on the node operation information of each node in the node operation information set: The quota threshold, measured occupancy, and minimum guaranteed value included in the node operation information are determined as the target quota threshold, target measured occupancy, and target minimum guaranteed value. In response to determining that the distributed node corresponding to the node operation information is a normal node, the target quota threshold is divided according to the target measured occupancy to obtain the node elastic quota corresponding to the normal node; In response to determining that the distributed node corresponding to the node operation information is the anchor node, the target quota threshold is divided according to the target measured occupancy and the target minimum guarantee value to obtain the node elastic quota corresponding to the anchor node; In response to determining that the distributed node corresponding to the node operation information is a designated node, the preset fixed quota is determined as the node elastic quota corresponding to the designated node; The node elasticity limit and the node operation information are determined as the initial quota information corresponding to the distributed node.

3. The method of claim 1, wherein, The step of performing steady-state redistribution processing on the constrained quota information set based on the supply node set to obtain the target quota information set includes: According to the preset buffer limit, the distributed node set corresponding to the quota information set within the constraint is subjected to buffer classification processing to obtain the set of points that need to be adjusted. Based on the set of required adjustment points and the set of supply nodes, the quota inter-adjustment processing is performed on the quota information set within the constraints to obtain the target quota information set.

4. The method of claim 1, wherein, The step of performing change determination processing on the distributed node set based on the target quota information set and the initial quota information set to obtain a quota change information set includes: For each distributed node in the distributed node set, perform the following change determination steps: The quota threshold included in the target quota information corresponding to the distributed node is determined as the quota threshold after scheduling, and the quota threshold included in the initial quota information corresponding to the distributed node is determined as the initial quota threshold. Based on the post-scheduling quota threshold and the initial quota threshold, the quota change information corresponding to the distributed node is determined.

5. The method of claim 1, wherein, The step of performing over-limit convergence processing on the initial quota information set based on the global quota upper limit to obtain the within-constraint quota information set includes: The initial quota information set is aligned with a carry to obtain the aligned quota information set. The distributed node set corresponding to the aligned quota information set is filtered and sorted to obtain the sequence of adjustment points to be adjusted. Based on the sequence of points to be adjusted and the aligned quota information set, the following adjustment steps are performed: The first point in the sequence of points to be adjusted is identified as the target point, and the target point is removed from the sequence of points to be adjusted to obtain the remaining sequence of points to be adjusted. The quota thresholds included in the aligned quota information corresponding to the target downward adjustment point are adjusted down to obtain the downward adjustment point quota information. The aligned quota information set and the aligned quota information corresponding to the target downward adjustment point are replaced with the downward adjustment point quota information to obtain the downward adjustment quota information set. In response to determining that the sum of the various quota thresholds included in the reduced quota information set is less than or equal to the global quota limit, the reduced quota information set is determined as the within-constraint quota information set; In response to determining that the sum of all quota thresholds included in the reduced quota information set is greater than the global quota limit, the reduced quota information set is determined as the aligned quota information set, and the remaining sequence of adjustment points to be reduced is determined as the sequence of adjustment points to be reduced, so as to execute the reduction step again.

6. A distributed node control device based on differentiated quota scheduling, characterized by, include: The acquisition unit is configured to, in response to receiving a quota scheduling request for a distributed node set, acquire a set of node running information and the global quota limit corresponding to the distributed node set. The partitioning unit is configured to perform elastic quota partitioning on the distributed node set based on the node operation information set to obtain an initial quota information set; The convergence unit is configured to, in response to determining that the initial quota information set satisfies a preset over-limit condition relative to the global quota upper limit, perform over-limit convergence processing on the initial quota information set according to the global quota upper limit to obtain an in-constraint quota information set; The identification unit is configured to perform occupancy trend identification processing on the quota information set within the constraints to obtain a supply node set; The redistribution unit is configured to perform steady-state redistribution processing on the constrained quota information set according to the supply node set to obtain the target quota information set; The determining unit is configured to perform change determination processing on the distributed node set based on the target quota information set and the initial quota information set to obtain a quota change information set; The control unit is configured to perform batch load control on the distributed node set based on the quota change information set and the target quota information set, wherein performing batch load control on the distributed node set based on the quota change information set and the target quota information set includes: performing batch division processing on the distributed node set based on the quota change information set to obtain a downgrade batch node set and an upgrade batch node set; performing load control on the downgrade batch node set based on the target quota information set; and performing load control on the upgrade batch node set based on the target quota information set in response to determining that the downgrade batch node set has completed load control.

7. An electronic device, comprising: include: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-5.

8. A computer readable medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-5.

Images