WO2023124347A1 - 一种存储节点的配置方法以及相关装置 - Google Patents

一种存储节点的配置方法以及相关装置 Download PDF

Info

Publication number
WO2023124347A1
WO2023124347A1 PCT/CN2022/123763 CN2022123763W WO2023124347A1 WO 2023124347 A1 WO2023124347 A1 WO 2023124347A1 CN 2022123763 W CN2022123763 W CN 2022123763W WO 2023124347 A1 WO2023124347 A1 WO 2023124347A1
Authority
WO
WIPO (PCT)
Prior art keywords
storage node
bandwidth
node
configuration information
target configuration
Prior art date
Application number
PCT/CN2022/123763
Other languages
English (en)
French (fr)
Inventor
徐洁
余洲
玄在雄
周建华
陈律
Original Assignee
华为云计算技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为云计算技术有限公司 filed Critical 华为云计算技术有限公司
Priority to EP22913612.2A priority Critical patent/EP4443833A1/en
Publication of WO2023124347A1 publication Critical patent/WO2023124347A1/zh
Priority to US18/759,328 priority patent/US20240354247A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
    • H04L41/5054—Automatic deployment of services triggered by the service manager, e.g. service implementation by automatic configuration of network components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/0223—User address space allocation, e.g. contiguous or non contiguous base addressing
    • G06F12/023—Free address space management
    • G06F12/0253—Garbage collection, i.e. reclamation of unreferenced memory
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0894—Policy-based network configuration management
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0896—Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
    • H04L41/5051—Service on demand, e.g. definition and deployment of services in real time
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
    • G06F2212/72—Details relating to flash memory management
    • G06F2212/7205—Cleaning, compaction, garbage collection, erase control
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Definitions

  • the embodiments of the present application relate to the field of computers, and in particular, to a storage node configuration method and a related device.
  • Block storage service is a cloud storage service provided by cloud service providers.
  • Block storage service can provide tenants with highly reliable and elastically expandable cloud hard disks.
  • the block storage service generally deploys distributed storage software on multiple storage nodes.
  • the distributed storage software virtualizes the storage hardware into a storage pool, and virtual cloud hard disks of different specifications can be separated from the storage pool.
  • Each storage node will be equipped with a large number of high-performance, large-capacity solid state drives (solid state drives, SSDs).
  • each storage node Since the access bandwidth provided by each storage node is limited, when the burst access bandwidth exceeds the maximum access bandwidth of the storage pool, the latency for tenants to access cloud disks will increase, and the storage node will limit the burst access bandwidth.
  • storage nodes use high-speed storage media as caches to absorb the sudden access bandwidth of tenants through the cache.
  • adding a high-speed storage medium to a storage node will increase the cost of the storage node. At the same time, it will cause a waste of storage resources when there are no sudden accesses from tenants.
  • Embodiments of the present application provide a method for configuring a storage node and a related device, which are used to reduce the cost of burst access to the storage node.
  • the embodiment of the present application provides a storage node configuration method, which can be executed by the management node, or by the components of the management node, such as the processor, chip or chip system of the management node, or by the Or a logical module or software implementation of part of the management node functions.
  • the configuration method provided by the embodiment of the present application includes: the management node obtains service demand information, and the service demand information includes the burst access time, burst access bandwidth, or burst access time of the computing node to the storage node in the future. Send access duration.
  • the management node determines the target configuration information according to the business demand information.
  • the target configuration information is used to configure the peak bandwidth and peak bandwidth duration of the storage node.
  • the peak bandwidth and peak bandwidth duration are related to the burst access time and burst Match the bandwidth or burst access duration, that is, the storage node configured according to the target configuration information can absorb the burst access bandwidth of the computing node to the storage node, and the management node sends the target configuration information to the storage node.
  • the management node can obtain the access service requirements of computing nodes and storage nodes, including burst access time, burst access bandwidth, or burst access duration, and determine the configuration information of storage nodes according to access service requirements, so that storage nodes It can increase the peak bandwidth to absorb the burst access bandwidth when the burst access arrives, which improves the ability of the storage node to cope with the burst access. Nodes deal with the cost of burst access, saving storage resources.
  • the target configuration information includes one or more of the following information: garbage collection GC waterline gear, power consumption gear, voltage gear, where the garbage collection waterline indicates the amount of data for garbage collection .
  • Configuring the GC waterline gear of the storage node can enable the storage node to perform garbage collection before the burst access time, so as to avoid garbage collection occupying the access bandwidth after the burst access time arrives.
  • Configuring the power consumption level of the storage node can make the storage node break through the power consumption limit for a short time, thereby temporarily increasing the peak bandwidth of the storage node to absorb the burst access bandwidth.
  • Configuring the voltage level of the storage node increases the write voltage to increase the write bandwidth of the computing node on the storage node.
  • the storage node in the embodiment of this application includes one or more solid-state disk SSDs, and the garbage collection GC waterline gear, power consumption gear or voltage gear of the storage node can be configured specifically for one or more SSDs in the storage node. Garbage collection GC waterline gear, power consumption gear or voltage gear.
  • the management node can configure the peak bandwidth and peak bandwidth duration of the storage node through multiple configuration modes, thereby improving the feasibility of the configuration solution.
  • the management node receives the business demand information sent by the computing node, that is, the computing node can set the burst access time, burst access bandwidth, burst access duration, etc.
  • the information is sent to the management node in advance.
  • the management node determines the predicted value of the service demand information based on the historical access information of the computing node to the storage node.
  • the historical access information includes the historical burst access time, historical burst access bandwidth, and historical burst access duration of the computing node to the management node.
  • the management node can predict and wait for business demand information based on historical access information.
  • the management node can obtain business demand information based on multiple methods, including direct acquisition from computing nodes and prediction through historical access information, so as to increase the diversity of ways to obtain business demand information in the solution.
  • the target configuration information is configured successfully on the storage node and starts counting the preparation time.
  • the management node sends a query command to the storage node.
  • the query command is used to query whether the storage node has reached the preparation completion time.
  • the preparation completion time is determined based on the burst access time and the preparation time.
  • the preparation time is the storage node’s The time required for the access bandwidth to reach the peak bandwidth.
  • the management node before sending the acceleration status command to the storage node, the management node will first send a query command to check whether the storage node is ready, which improves the reliability of the storage node to accelerate to the peak bandwidth.
  • the management node monitors the access bandwidth of the storage node. When the access bandwidth is less than the threshold or reaches the peak bandwidth duration, the management node sends an exit acceleration state command to the storage node, and the exit acceleration state command is used to restore the peak bandwidth to the normal bandwidth at the storage node.
  • the management section monitors the access bandwidth of the storage node in real time, and at the same time, there are various trigger conditions for the storage node to exit the accelerated state, including the burst access bandwidth being less than the threshold or the duration of the peak bandwidth, so that the storage node can exit the accelerated state in time, In this way, the waste of peak bandwidth resources is avoided, and the service life of the storage nodes is avoided at the same time.
  • the management node determines the target peak bandwidth and the target peak bandwidth duration from the configuration table according to the service requirement information, and the configuration table includes different Configure the corresponding peak bandwidth, peak bandwidth duration or preparation time, and the management node determines the target configuration information according to the target peak bandwidth and target peak bandwidth duration, that is, selects the corresponding GC waterline gear according to the target peak bandwidth and target peak bandwidth duration , power consumption gear and voltage gear.
  • the management node can directly select the corresponding peak bandwidth and peak bandwidth duration from the configuration table based on the service requirement information, which improves the efficiency of task configuration.
  • the second aspect of the embodiment of the present application provides a method for configuring a storage node.
  • the method may be executed by the storage node, or by a component of the storage node, such as a processor, a chip, or a chip system of the storage node, or by A logic module or software implementation that can realize all or part of the storage node functions.
  • the configuration method provided by the embodiment of the present application includes: the storage node receives target configuration information sent by the management node, and the target configuration information is used to configure the peak bandwidth and peak bandwidth duration of the storage node.
  • the storage node configures the storage node according to the target configuration information, so that the storage node handles the burst access of the computing node to the storage node.
  • the management node can obtain the access service requirements of the computing nodes and storage nodes, and determine the configuration information of the storage nodes according to the access service requirements, so that the storage nodes can absorb the burst access bandwidth when the burst access arrives, which improves the storage node The ability to deal with burst access. At the same time, the storage node does not increase the storage medium to absorb the burst access bandwidth, saving storage resources.
  • the target configuration information includes one or more of the following information: garbage collection GC waterline gear, power consumption gear or voltage gear.
  • Configuring the GC waterline gear of the storage node can enable the storage node to perform garbage collection before the burst access time, so as to avoid garbage collection occupying the access bandwidth after the burst access time arrives.
  • Configuring the power consumption level of the storage node can make the storage node break through the power consumption limit for a short time, thereby temporarily increasing the peak bandwidth of the storage node to absorb the burst access bandwidth.
  • Configuring the voltage level of the storage node for example, configuring the write voltage level of the storage node, increases the write voltage to increase the write bandwidth of the computing node on the storage node.
  • the peak bandwidth and the peak bandwidth duration of the storage node can be configured in various configuration manners, thereby improving the feasibility of the configuration solution.
  • the storage node receives the query instruction sent by the management node, and the query instruction is used to query whether the storage node has reached the preparation completion time.
  • the preparation completion time is determined based on the burst access time and the preparation time, and the preparation time is the storage node's The time required for the access bandwidth to reach the peak bandwidth. If the storage node arrives at the preparation completion time, the storage node sends a query command response to the management node, and the query command response indicates that the preparation time corresponding to the target configuration information of the storage node has been ready.
  • the query command sent by the management node will be received, and a query command response will be sent to the management node.
  • the query command response can indicate whether the storage node is ready, thereby improving the reliability of the storage node to accelerate to the peak bandwidth.
  • the storage node receives the acceleration status command sent by the management node, and the acceleration status command is used to start the peak bandwidth on the storage node, that is, the acceleration status command can make the peak bandwidth and peak bandwidth duration corresponding to the target configuration information take effect .
  • the storage node increases the access bandwidth to the peak bandwidth after receiving the acceleration state command sent by the management node, which avoids waste of peak bandwidth resources.
  • the storage node configures the GC waterline gear of the storage node according to the target configuration information, and increases the amount of time the storage node uses for garbage processing before the burst access time. bandwidth.
  • configuring the GC waterline position of the storage node can enable the storage node to perform garbage collection before the burst access time, so as to prevent the garbage collection from occupying the access bandwidth after the burst access time arrives, so that when the burst access arrives Increased bandwidth resources for handling burst access.
  • the storage node configures the power consumption gear of the storage node according to the target configuration information, so that the power consumption threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • configuring the power consumption level of the storage node can make the storage node break through the power consumption limit in a short time, thereby increasing the peak bandwidth of the storage node in a short time to absorb the burst access bandwidth, thereby improving the storage node's response to burst access. ability.
  • the storage node configures the voltage level of the storage node according to the target configuration information, so that the voltage threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • the voltage level of the storage node is configured to increase the access bandwidth of the storage node, thereby improving the ability of the storage node to cope with burst access.
  • a third aspect of the embodiments of the present application provides a management node, where the management node includes a transceiver unit and a processing unit.
  • the transceiver unit is used to obtain service requirement information, and the service requirement information includes burst access time, burst access bandwidth, or burst access duration of the computing node to the storage node.
  • the processing unit is configured to determine target configuration information according to service requirement information, and the target configuration information is used to configure peak bandwidth and peak bandwidth duration of the storage node.
  • the transceiver unit is also configured to send target configuration information to the storage node.
  • the target configuration information includes one or more items of information: GC waterline gear, power consumption gear, and voltage gear.
  • the transceiver unit is specifically configured to receive service requirement information sent by the computing node.
  • the processing unit is further configured to determine service requirement information according to historical access information of the computing node to the storage node.
  • the transceiver unit when the burst access time is reached, the transceiver unit is also used to send a query instruction to the storage node.
  • the query instruction is used to query whether the storage node has reached the preparation completion time, and the preparation completion time is based on the burst access time and the preparation time. The time is determined, and the preparation time is the time required for the access bandwidth of the storage node to reach the peak bandwidth.
  • the transceiver unit is further configured to send an acceleration state command to the storage node, and the acceleration state command is used to start the peak bandwidth on the storage node.
  • the processing unit is further configured to monitor the access bandwidth of the storage node.
  • the transceiver unit is further configured to send an exit acceleration state command to the storage node, and the exit acceleration state command is used to restore the peak bandwidth to a normal bandwidth at the storage node.
  • the processing unit is specifically configured to determine the target peak bandwidth and target peak bandwidth duration from a configuration table according to service demand information, and the configuration table includes peak bandwidth, peak bandwidth duration or Preparation time.
  • the processing unit is specifically configured to determine the target configuration information according to the target peak bandwidth and the target peak bandwidth duration.
  • a fourth aspect of the embodiments of the present application provides a storage node, where the storage node includes a transceiver unit and a processing unit.
  • the transceiver unit is used for receiving the target configuration information sent by the management node, and the target configuration information is used for configuring the peak bandwidth and peak bandwidth duration of the storage node.
  • the processing unit is configured to configure the storage node according to the target configuration information, so that the storage node processes the burst access of the computing node to the storage node.
  • the target configuration information includes one or more of the following information: garbage collection GC waterline gear, power consumption gear or voltage gear.
  • the transceiver unit is further configured to receive an acceleration status command sent by the management node, where the acceleration status command is used to start the peak bandwidth on the storage node.
  • the processing unit is specifically configured to configure the GC pipeline gear of the SSD in the storage node according to the target configuration information, so as to increase the bandwidth of the storage node for garbage processing before the burst access time.
  • the processing unit is specifically configured to configure the power consumption gear of the SSD in the storage node according to the target configuration information, so that the power consumption threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • the processing unit is specifically configured to configure the voltage gear of the SSD in the storage node according to the target configuration information, so that the voltage threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • a fifth aspect of the embodiments of the present application provides a storage node configuration system, where the configuration system includes a management node and a storage node.
  • the management node is configured to execute the method described in the first aspect or any possible implementation manner of the first aspect
  • the storage node is configured to execute the method described in the second aspect or any possible implementation manner of the second aspect .
  • the fifth aspect of the embodiment of the present application provides a configuration device, including a processor, the processor is coupled with the memory, and the processor is used to store instructions.
  • the configuration device performs the above-mentioned first aspect or the first The method described in any possible implementation manner of the second aspect, or the configuration device executes the method described in the second aspect or any possible implementation manner of the second aspect.
  • the sixth aspect of the embodiments of the present application provides a computer-readable storage medium on which instructions are stored.
  • the computer executes the above-mentioned first aspect or any one of the possible implementation manners of the first aspect. method, or to cause a computer to execute the method described in the second aspect or any possible implementation manner of the second aspect.
  • the sixth aspect of the embodiment of the present application provides a computer program product, the computer program product includes instructions, and it is characterized in that, when the instructions are executed, the computer can realize the above-mentioned first aspect or any possible implementation manner of the first aspect The method described above, or to enable a computer to implement the method described in the above second aspect or any possible implementation manner of the second aspect.
  • beneficial effects that can be achieved by any management node, storage node, configuration system, configuration device, computer readable medium, or computer program product provided above can refer to the beneficial effects in the corresponding method, which will not be repeated here. repeat.
  • FIG. 1 is a schematic diagram of a system architecture of a storage system provided by an embodiment of the present application
  • FIG. 2 is a schematic diagram of a method for configuring a storage node provided in an embodiment of the present application
  • FIG. 3 is a schematic flowchart of a method for configuring a storage node provided in an embodiment of the present application
  • FIG. 4 is a schematic diagram of another method for configuring a storage node provided in an embodiment of the present application.
  • FIG. 5 is a schematic diagram of changes in peak bandwidth of a storage node provided in an embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of a configuration device provided by an embodiment of the present application.
  • FIG. 7 is a schematic structural diagram of a configuration device provided in an embodiment of the present application.
  • FIG. 8 is a schematic structural diagram of a configuration system provided by an embodiment of the present application.
  • Embodiments of the present application provide a task scheduling method and a task scheduling device, which are used to improve the efficiency of task scheduling.
  • words such as “exemplary” or “for example” are used as examples, illustrations or illustrations. Any embodiment or design scheme described as “exemplary” or “for example” in the embodiments of the present application shall not be interpreted as being more preferred or more advantageous than other embodiments or design schemes. Rather, the use of words such as “exemplary” or “such as” is intended to present related concepts in a concrete manner.
  • GC Garbage collection
  • Solid state disk refers to a hard disk made of an array of solid-state electronic memory chips.
  • solid state drives There are two types of storage media for solid state drives, one is to use flash memory (flash) as the storage medium, and the other is to use dynamic random access memory (DRAM) as the storage medium.
  • flash flash memory
  • DRAM dynamic random access memory
  • Small computer system interface is a system-level interface standard used between computers and peripheral devices, such as hard disks, floppy drives, optical drives, printers or scanners.
  • a hard disk using SCSI is also called a SCSI hard disk.
  • iSCSI Internet Small Computer System Interface
  • FIG. 1 is a schematic diagram of a system architecture of a storage system provided by an embodiment of the present application.
  • the storage system 100 includes a storage node 101, a management node 102, and a computing node 103.
  • the storage node 101 includes a solid state disk SSD configuration module 1011 and one or more SSD1012, wherein the SSD configuration module 1011 is configured to configure the peak bandwidth and peak duration of one or more SSD1012 according to the configuration information sent by the management node 102, configured
  • the specific content includes garbage collection GC waterline stalls, power consumption stalls or voltage stalls.
  • multiple SSDs 1012 are virtualized into storage pools through distributed software, and the storage pools can divide virtual cloud disks of different capacities and provide them to the computing nodes 103 for use.
  • tenant applications are deployed on computing nodes 103, and the tenant applications use virtual cloud disks as data disks or system disks.
  • the management node 102 is used to manage the processes in the storage node 101 and the computing node 103, and provide functions such as alarm, monitoring, log, or configuration.
  • the management node 102 includes an SSD acceleration configuration module 1021, and the SSD acceleration configuration module 1021 is configured to determine configuration information of the storage node 101 according to service requirement information.
  • the management node 102 communicates with the storage node 101 through a small computer system interface SCSI or an Internet small computer system interface ISCSI.
  • the storage node 101, the management node 102, and the computing node 103 may be respectively deployed on different physical devices, and the management node 102 may also be deployed on a virtual machine, which is not specifically limited.
  • the storage node 101 and the management node 102 can also be integrated in the same device.
  • the device can be based on a The integrated hardware modules implement the functions of the storage node 101 and the management node 102 described above.
  • FIG. 2 is a schematic flowchart of a configuration method provided by an embodiment of the present application.
  • the storage node configuration method provided by the embodiment of the present application includes the following steps:
  • the management node obtains service requirement information, and the service requirement information includes the burst access time, burst access bandwidth, or burst access duration of the computing node to the storage node.
  • the management node obtains service requirement information, and the service requirement information includes the burst access time, burst access bandwidth, or burst access duration of the computing node to the storage node.
  • the service demand information may be multiple burst access times, burst access bandwidth, and burst access duration of the storage node within a period of time in the future.
  • the storage node has two burst accesses during the period from 12:00 to 18:00 of the next day, and the two burst accesses correspond to different burst access times, burst access bandwidths, and burst access durations.
  • the management node can obtain business demand information in various ways, for example, the management node can receive the business demand information sent by the computing node; or the management node can obtain the business demand information of the computing node by performing data mining on the computing node; or The management node obtains the predicted value of the service demand information according to the historical access information of the computing node to the storage node. Specifically, the management node predicts the burst access time, Burst access bandwidth and burst access duration.
  • burst access in the embodiment of the present application includes burst write and burst read. Unless otherwise specified, the burst access in the embodiment of the present application is described using burst write as an example.
  • FIG. 3 is a schematic flowchart of a method for configuring a storage node provided in an embodiment of the present application.
  • the management node predicts and obtains service demand information based on the historical access information of the computing node. Specifically, the management node predicts the burst access time, burst access bandwidth W, and burst Visit duration T.
  • the management node determines target configuration information according to the service requirement information, and the target configuration information is used to configure peak bandwidth and peak bandwidth duration of the storage node.
  • the management node determines the target configuration information according to the service requirement information, and the target configuration information includes one or more of the following: GC waterline gear, power consumption gear, and voltage gear. Specifically, the management node determines the target peak bandwidth and the target peak bandwidth duration corresponding to the business requirement information from the configuration table according to the service requirement information, and selects the garbage collection GC waterline file corresponding to the target peak bandwidth and the target peak bandwidth duration bit, power consumption level or voltage level as the target configuration information.
  • the storage node includes one or more SSDs
  • the target configuration information sent by the management node to the storage node is specifically used to indicate the garbage collection GC waterline stalls and function of one or more SSDs in the storage node. Consumption gear and voltage gear.
  • Table 1 is a configuration table provided by the embodiment of the present application.
  • the configuration table includes GC waterline gear, power consumption gear, write voltage gear, peak write bandwidth, peak write bandwidth duration or preparation time. It can be seen from Table 1 that the higher the peak write bandwidth of the storage node, the longer the preparation time required by the storage node, and the longer the peak write bandwidth duration of the storage node, the longer the preparation time required by the storage node.
  • the storage node collects garbage in advance during the preparation time, adjusts the working mode of the storage node, or performs data movement inside the storage node.
  • the GC waterline gears of the storage node include three gears of 0, 1, and 2, and the power consumption gear and the write voltage gear include two gears of 0 and 1.
  • the higher the write voltage level the higher the write voltage threshold of the storage node.
  • the configuration table can be provided to the computing nodes through a user manual, so that users or cloud vendors can easily query the preparation time.
  • the management node determines the appropriate peak write bandwidth and peak write bandwidth duration in the configuration table according to the burst write bandwidth and burst write bandwidth duration in the service requirement information, and The garbage collection GC waterline gear, power consumption gear or write voltage gear corresponding to the peak write bandwidth and the peak write bandwidth duration are used as the target configuration information.
  • the following describes the specific process for the management node to determine the target configuration information in the configuration table in the embodiment of the present application.
  • FIG. 4 is a schematic diagram of determining target configuration information provided by an embodiment of the present application.
  • the management node After the management node obtains the business demand information, it reads the middle peak bandwidth and peak bandwidth duration of the configuration table, assuming that the business demand information burst access bandwidth is W, and the burst access duration is T.
  • the peak bandwidth read in the table is WS, and the peak bandwidth duration is TS.
  • the minimum configuration of WS and TS is selected as the target configuration information among all WS and TS satisfying the conditions.
  • WS and TS exist in the configuration table and satisfy WS ⁇ W and TS ⁇ T, then select the configuration with the largest WS among all configurations that meet the conditions, if there is TS such that WS*TS ⁇ W*T, then Select the configuration with the smallest TS among the configurations that meet the conditions. If there is no TS such that WS*TS ⁇ W*T, then TS selects the largest configuration.
  • the configuration with the largest WS and TS is selected as the target configuration information.
  • the management node determines the target configuration information from the configuration table.
  • the configuration table takes Table 1 as an example.
  • the configuration satisfying WS*TS ⁇ W*T includes state 9 and state 12, where WS in state 9 ⁇ WS in state 12, so the configuration corresponding to state 9 is selected as the target configuration information.
  • the management node determines target configuration information according to service requirement information.
  • the business requirement information indicates that the burst access time is 12 noon, the burst write bandwidth is 5000MB/s, and the duration is 60s.
  • the management node queries the peak write bandwidth and peak write bandwidth duration in the configuration table.
  • the configuration table takes Table 1 as an example. In all states of the SSD, the peak write bandwidth is greater than or equal to 5000MB/s and the peak write bandwidth duration is greater than or equal to 60s.
  • the options include state 5, state 6, state 8, state 9, state 11, and state 12.
  • state 5 has the smallest amount of data, and the amount of data is equal to the peak write bandwidth and the duration of the peak write bandwidth. Therefore, the management node takes the GC waterline gear, power consumption gear and voltage gear corresponding to state 5 as the target configuration information.
  • the management node sends target configuration information to the storage node.
  • the management node sends the target configuration information to the storage node. Specifically, the management node sends one or more of the following information in the target configuration information to the storage node through the SSD peak performance configuration interface: GC waterline gear, power consumption gear, voltage range.
  • the storage node can enter the ready state after receiving the target configuration information sent by the management node, or the storage node can enter the ready state after receiving the preparation instruction sent by the management node, which is not specifically limited.
  • the storage node in the ready state starts to count the preparation time.
  • the management node sends target configuration information to the storage node.
  • the target configuration information is the GC waterline gear, power consumption gear and voltage gear corresponding to state 5 in Table 1, and the management node sends the configuration corresponding to state 5 to the storage node.
  • the storage node After the storage node receives the configuration corresponding to state 5 in Table 1, since the preparation time required to enter state 5 is 3600s, the storage node needs to enter the preparation state of state 5 at least 3600s in advance.
  • the storage node enters the ready state after receiving the preparation command sent by the management node as an example, assuming that the burst access time is 12:00, at 10:30, the management node sends Send the prepare command to enter state 5.
  • the embodiment of the present application provides a storage node with configurable peak bandwidth, and the storage node provides an SSD peak performance configuration interface, so that the management node calls the SSD peak performance configuration interface according to the target configuration information to change the storage peak write bandwidth.
  • the storage node configures the storage node according to the target configuration information.
  • the storage node configures the storage node according to the target configuration information, including configuring the GC waterline gear, power consumption gear, and voltage gear of the storage node.
  • the configuration methods of the above storage nodes can be configured in time-sharing, or can be superimposed at the same time. No limit. The above configuration methods are described in detail below.
  • the storage node When the storage node configures the GC waterline position of the storage node according to the target configuration information, the storage node starts garbage collection in advance of the preparation time, so as to avoid the garbage collection from occupying the access bandwidth after the burst access time arrives. Specifically, the storage node reserves a part of blank block space for absorbing the burst load, and the amount of data to absorb the burst load is determined by the GC waterline gear.
  • a blank block refers to a storage space that does not store data, and a block is a unit of data storage and processing. When the burst load comes, the blank block can absorb the burst write bandwidth at a speed close to the normal write bandwidth until the space consumption of the blank block is completed, and the sequential write bandwidth is, for example, 4.5GB/S.
  • the storage node When the storage node configures the power consumption level of the storage node according to the target configuration information, the storage node increases the power consumption level of the storage node according to the target configuration information, so that the storage node breaks through the power consumption limit in a short time, thereby increasing the power consumption of the storage node in a short time. Peak bandwidth absorbs burst access bandwidth.
  • the power consumption threshold of the SSD of the storage node in the normal state is 25W.
  • the storage node increases the power consumption threshold of the SSD to 40W according to the target configuration information.
  • the write bandwidth of the SSD of the storage node is fully released, and some bandwidth of the storage node can be increased from 4.5GB/S increased to 6GB/S.
  • the storage node can adjust the write voltage threshold of the flash memory particles of the SSD.
  • the higher the write voltage threshold the higher the write voltage step, so that The fewer steps for the SSD of the storage node to reach the specified state, the higher the write bandwidth.
  • the high writing voltage level can increase the writing bandwidth by 50%.
  • the management node sends a query command to the storage node, and the query command is used to query whether the storage node is ready to arrive. Time, that is, to query whether the preparation time of the storage node has been counted. If the preparation time has been completed, the management node determines that the preparation of the storage node has been completed. Specifically, the storage node may send a query command response to the management node, where the query command response is used to indicate the preparation status of the storage node to the management node.
  • the storage node After the management node determines that the storage node has reached the ready completion time, the storage node sends an acceleration status command to the management node.
  • the acceleration status command is used to set the storage node to enable the acceleration status.
  • the access bandwidth of the storage node reaches the peak bandwidth.
  • the management node monitors the access bandwidth of the storage node in real time. When the burst access bandwidth of the storage node ends, or the duration of the peak bandwidth has passed, the management node sends an exit acceleration state command to the storage node, and the storage node receives the exit acceleration state command. , the bandwidth of the storage node recovers from the peak bandwidth to the normal bandwidth.
  • the storage node After the storage node receives the target configuration information sent by the management node, the storage node enters a ready state.
  • the management node sends a query command to the storage node to check whether the storage node is ready. If the storage node is ready, the management node sends an acceleration status command to the storage node. If the preparation is not completed, the storage node is not successfully configured according to the target configuration information, and the storage node does not have the peak adjustment capability.
  • the management node continuously monitors the access bandwidth of the storage node. If the burst access ends or the duration of the peak bandwidth has passed, the management node sends an exit acceleration state command to the storage node, so that the storage node is Bandwidth returns to normal bandwidth.
  • FIG. 5 is a schematic diagram of changes in peak write bandwidth of a storage node provided in an embodiment of the present application. As shown in Figure 5, after the peak write bandwidth of the storage node is successfully configured according to the target configuration information, when the storage node receives the acceleration state command, the peak bandwidth of the storage node will increase, and when the storage node receives the exit acceleration state command, the storage node The peak bandwidth of the node is reduced to the normal value of the peak bandwidth.
  • the conventional peak write bandwidth value of the SSD of the storage node is W0, and the target configuration information is sent at Ts time, and the peak write bandwidth of the SSD is configured as W1 according to the target configuration information, and lasts for T1 time.
  • the preparation time corresponding to W1 and T1 is (T0-Ts). After the (T0-Ts) preparation time passes, at time T0, the peak write bandwidth of the SSD reaches W1, and the sustainable time is T1.
  • the peak write bandwidth of SSD is configured as W2, and the duration is T2, and the preparation time corresponding to W2 and T2 is (T0'-Ts'), after (T0'-Ts') preparation time, SSD The peak write bandwidth reaches W2 and lasts for T2.
  • the management node can obtain the access service requirements of the computing nodes, and determine the configuration information of the storage nodes according to the access service requirements, so that the storage nodes can absorb the burst access bandwidth when the burst access arrives, which improves the response of the storage node to bursts.
  • the storage node does not add a new storage medium to absorb the burst access bandwidth, which reduces the cost of the storage node to deal with burst access and saves storage resources.
  • FIG. 6 is a schematic structural diagram of a configuration device provided by an embodiment of the present application.
  • the apparatus is used to implement various steps performed by the management node or the storage node in the foregoing embodiments.
  • the configuration apparatus 600 includes a transceiver unit 601 and a processing unit 602 .
  • the configuration device 600 is used to implement the method performed by the management node in the above method embodiment, that is, the management node may include a transceiver unit 601 and a processing unit 602 .
  • the transceiver unit 601 is used to obtain service requirement information, and the service requirement information includes burst access time, burst access bandwidth, or burst access duration of computing nodes to storage nodes.
  • the processing unit 602 is configured to determine target configuration information according to service requirement information, and the target configuration information is used to configure peak bandwidth and peak bandwidth duration of the storage node.
  • the transceiver unit is also configured to send target configuration information to the storage node.
  • the target configuration information includes one or more items of information: garbage collection GC waterline gear, power consumption gear, voltage gear, wherein the garbage collection waterline indicates the amount of data for garbage collection .
  • the transceiver unit 601 is specifically configured to receive service requirement information sent by the computing node.
  • the processing unit is further configured to determine the predicted value of the service requirement information according to historical access information of the computing node to the storage node.
  • the transceiver unit 601 when the burst access time is reached, the transceiver unit 601 is also configured to send a query instruction to the storage node, the query instruction is used to query whether the storage node has reached the preparation completion time, and the preparation completion time is based on the burst access time and The preparation time is determined, and the preparation time is the time required for the access bandwidth of the storage node to reach the peak bandwidth.
  • the transceiver unit 601 is further configured to send an acceleration state command to the storage node, where the acceleration state command is used to start the peak bandwidth on the storage node.
  • the processing unit 602 is also configured to monitor the access bandwidth of the storage node. When the access bandwidth is less than the threshold or reaches the peak bandwidth duration, the transceiver unit 601 is further configured to send an exit acceleration state command to the storage node, and the exit acceleration state command is used to restore the peak bandwidth to the normal bandwidth at the storage node.
  • the processing unit 602 is specifically configured to determine the target peak bandwidth and the target peak bandwidth duration from the configuration table according to the service demand information, and the configuration table includes the peak bandwidth and the peak bandwidth duration corresponding to different configurations of the storage nodes. or prep time.
  • the processing unit 602 is specifically configured to determine target configuration information according to the target peak bandwidth and the target peak bandwidth duration.
  • the configuration apparatus 600 is configured to implement the method performed by the storage node in the foregoing method embodiments, that is, the storage node may include a transceiver unit 601 and a processing unit 602 .
  • the transceiver unit 601 is configured to receive target configuration information sent by the management node, and the target configuration information is used to configure the peak bandwidth and peak bandwidth duration of the storage node.
  • the processing unit 602 is configured to configure the storage node according to the target configuration information, so that the storage node processes the burst access of the computing node to the storage node.
  • the target configuration information includes one or more of the following information: garbage collection GC waterline gear, power consumption gear or voltage gear, wherein the garbage collection waterline indicates the amount of data for garbage collection .
  • the transceiver unit 601 is further configured to receive an acceleration status command sent by the management node, where the acceleration status command is used to start the peak bandwidth on the storage node.
  • the processing unit 602 is specifically configured to configure the GC pipeline gear of the SSD in the storage node according to the target configuration information, and increase the bandwidth of the storage node for garbage processing before the burst access time.
  • the processing unit 602 is specifically configured to configure the power consumption gear of the SSD in the storage node according to the target configuration information, so that the power consumption threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • the processing unit 602 is specifically configured to configure the voltage level of the SSD in the storage node according to the target configuration information, so that the voltage threshold of the storage node increases, thereby increasing the bandwidth of the storage node.
  • each unit in the device can be implemented in the form of software called by the processing element; they can also be implemented in the form of hardware; some units can also be implemented in the form of software called by the processing element, and some units can be implemented in the form of hardware.
  • each unit can be a separate processing element, or can be integrated in a certain chip of the device.
  • it can also be stored in the memory in the form of a program, which is called and executed by a certain processing element of the device. Function.
  • all or part of these units can be integrated together, or implemented independently.
  • the processing element mentioned here may also be a processor, which may be an integrated circuit with signal processing capability.
  • each step of the above method or each unit above may be implemented by an integrated logic circuit of hardware in the processor element or implemented in the form of software called by the processing element.
  • FIG. 7 is a schematic diagram of a configuration device provided by an embodiment of the present application.
  • the configuration device 700 includes: a processor 710 , a memory 720 and an interface 730 , and the processor 710 , the memory 720 and the interface 730 are coupled through a bus (not marked in the figure).
  • the memory 720 stores instructions, and when the execution instructions in the memory 720 are executed, the configuration device 700 executes the method executed by the management node or the storage node in the foregoing method embodiments.
  • Configuration device 700 may be one or more integrated circuits configured to implement the above method, for example: one or more specific integrated circuits (application specific integrated circuit, ASIC), or, one or more microprocessors (digital signal processor , DSP), or, one or more field programmable gate arrays (field programmable gate array, FPGA), or a combination of at least two of these integrated circuit forms.
  • ASIC application specific integrated circuit
  • DSP digital signal processor
  • FPGA field programmable gate array
  • the units in the device can be implemented in the form of a processing element scheduler
  • the processing element can be a general-purpose processor, such as a central processing unit (central processing unit, CPU) or other processors that can call programs.
  • CPU central processing unit
  • these units can be integrated together and implemented in the form of a system-on-a-chip (SOC).
  • Processor 710 may be a central processing unit (central processing unit, CPU), or other general-purpose processors, digital signal processors (digital signal processor, DSP), application specific integrated circuits (application specific integrated circuits, ASICs), on-site Programmable gate array (field programmable gate array, FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof.
  • a general-purpose processor can be a microprocessor, or any conventional processor.
  • the memory 720 may include read-only memory and random-access memory, and provides instructions and data to the processor 710 .
  • Memory 720 may also include non-volatile random access memory.
  • the memory 720 may be provided with multiple partitions, each of which is used to store private keys of different software modules.
  • Memory 720 can be volatile memory or nonvolatile memory, or can include both volatile and nonvolatile memory.
  • the non-volatile memory can be read-only memory (read-only memory, ROM), programmable read-only memory (programmable ROM, PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrically programmable Erases programmable read-only memory (electrically EPROM, EEPROM) or flash memory.
  • Volatile memory can be random access memory (RAM), which acts as external cache memory.
  • RAM random access memory
  • SRAM static random access memory
  • DRAM dynamic random access memory
  • SDRAM synchronous dynamic random access memory
  • Double data rate synchronous dynamic random access memory double data date SDRAM, DDR SDRAM
  • enhanced SDRAM enhanced synchronous dynamic random access memory
  • SLDRAM synchronous connection dynamic random access memory
  • direct rambus RAM direct rambus RAM
  • the bus may also include a power bus, a control bus, and a status signal bus.
  • the bus can be a peripheral component interconnect express (PCIe) bus, or an extended industry standard architecture (EISA) bus, a unified bus (Ubus or UB), a computer fast link (compute express link, CXL), cache coherent interconnect for accelerators (CCIX), etc.
  • PCIe peripheral component interconnect express
  • EISA extended industry standard architecture
  • Ubus or UB unified bus
  • CXL computer fast link
  • CXL cache coherent interconnect for accelerators
  • the bus can be divided into address bus, data bus, control bus and so on.
  • FIG. 8 is a schematic diagram of a configuration system provided by an embodiment of the present application.
  • the configuration system 800 includes a management node 801 and a storage node 802.
  • the management node 801 may be the management node in the above method embodiment, and the storage node 802 It may be the storage node in the foregoing method embodiments.
  • a computer-readable storage medium is also provided.
  • Computer-executable instructions are stored in the computer-readable storage medium.
  • the processor of the device executes the computer-executable instructions
  • the device executes the above-mentioned method embodiment.
  • a computer program product in another embodiment of the present application, includes computer-executable instructions, and the computer-executable instructions are stored in a computer-readable storage medium.
  • the processor of the device executes the computer-executed instruction, the device executes the method performed by the management node or the storage node in the foregoing method embodiments.
  • the disclosed system, device and method can be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.
  • the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.
  • the technical solution of the present application is essentially or part of the contribution to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), magnetic disk or optical disc, etc., which can store program codes. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Logic Circuits (AREA)

Abstract

本申请实施例公开了一种存储节点的配置方法以及相关装置，用于降低存储节点应对突发访问的成本。本申请实施例方法包括：获取业务需求信息，业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间，根据业务需求信息确定目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间，向存储节点发送目标配置信息。

Description

一种存储节点的配置方法以及相关装置
本申请要求于2021年12月31日提交中国专利局、申请号为“202111679415.2”、申请名称为“一种存储节点的配置方法以及相关装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及计算机领域，尤其涉及一种存储节点的配置方法以及相关装置。
背景技术
块存储服务是云服务厂商提供的一种云存储服务，块存储服务可为租户提供高可靠、可弹性扩容的云硬盘。块存储服务一般会在多个存储节点上部署分布式存储软件，分布式存储软件会将存储硬件虚拟成为存储池，存储池上可以分出不同规格的虚拟云硬盘。每个存储节点上会配备大量高性能、大容量的固态硬盘(solid state drive，SSD)。
由于每个存储节点能够提供的访问带宽有限，当突发访问带宽超过存储池的最大访问带宽时，租户访问云硬盘的延迟会增大，同时存储节点会限制突发访问带宽。目前为了应对租户的突发访问，存储节点通过增加高速存储介质作为缓存，通过缓存吸收租户的突发访问带宽。但是，存储节点增加高速存储介质会造成存储节点的成本增加，同时，在没有租户突发访问的时候又造成存储资源的浪费。
发明内容
本申请实施例提供了一种存储节点的配置方法以及相关装置，用于降低存储节点应突发访问的成本。
本申请实施例提供了一种存储节点的配置方法，该方法可以由管理节点执行，也可以由管理节点的部件，例如管理节点的处理器、芯片或芯片系统等执行，还可以由能实现全部或部分管理节点功能的逻辑模块或软件实现。以管理节点执行为例，本申请实施例提供的配置方法包括：管理节点获取业务需求信息，业务需求信息包括计算节点对存储节点的未来一段时间内的突发访问时刻、突发访问带宽或突发访问持续时间。管理节点根据业务需求信息确定目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间，该峰值带宽和峰值带宽持续时间与计算节点的对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间相匹配，即根据目标配置信息配置的存储节点能够吸收计算节点对存储节点的突发访问带宽，管理节点向存储节点发送目标配置信息。
本申请实施例中管理节点能够获取计算节点存储节点的访问业务需求，包括突发访问时刻、突发访问带宽或突发访问持续时间，并根据访问业务需求确定存储节点的配置信息，使得存储节点能够在突发访问到来时刻提高峰值带宽吸收突发访问带宽，提升了存储节点应对突发访问的能力，同时，存储节点没有增加新的存储介质吸收突发访问带宽，相比现有技术降低存储节点应对突发访问的成本，节约了存储资源。
一种可能的实施方式中，目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位、电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。配置存储节点的GC水线档位能够使得存储节点在突发访问时刻之前进行垃圾回收，从而避免在突发访问时间到来之后垃圾回收占用访问带宽。配置存储节点的功耗档位能够使得存储节点短时突破功耗限制，从而短时提升存储节点的峰值带宽以吸收突发访问带宽。配置存储节点的电压档位例如配置存储节点的写入电压档位，提升写入电压从而提升计算节点在存储节点的写带宽。本申请实施例中的存储节点包括一个或多个固态硬盘SSD，配置存储节点的垃圾回收GC水线档位、功耗档位或电压档位具体可以是配置存储节点中一个或多个SSD的垃圾回收GC水线档位、功耗档位或电压档位。
本申请实施例中管理节点能够通过多种配置方式配置存储节点的峰值带宽和峰值带宽持续时间，从而提升了配置方案的可实现性。
一种可能的实施方式中，管理节点获取业务需求信息的过程中，管理节点接收计算节点发送的业务需求信息，即计算节点能够将突发访问时刻、突发访问带宽和突发访问持续时间等信息提前发送给管理节点。或者，管理节点根据计算节点对存储节点的历史访问信息确定业务需求信息的预测值，历史访问信息包括计算节点对管理节点的历史突发访问时刻、历史突发访问带宽和历史突发访问持续时间，管理节点能够根据历史访问信息预测等到业务需求信息。
本申请实施例中管理节点能够基于多种方式获取业务需求信息，包括从计算节点直接获取的方式和通过历史访问信息预测的方式，从而提升方案中获取业务需求信息方式的多样性。
一种可能的实施方式中，管理节点向存储节点发送目标配置信息之后，目标配置信息在存储节点配置成功并开始准备时间计时。当存储节点到达突发访问时刻，管理节点向存储节点发送查询指令，查询指令用于查询存储节点是否到准备完成时刻，准备完成时刻基于突发访问时刻和准备时间确定，准备时间为存储节点的访问带宽到达峰值带宽的所需要的时间。当存储节点到达准备完成时刻，则向存储节点发送加速状态命令，加速状态命令用于在存储节点启动峰值带宽。
本申请实施例中管理节点在向存储节点发送加速状态命令之前，会先发送查询指令查询存储节点是否准备完成，从提升了存储节点加速到达峰值带宽的可靠性。
一种可能的实施方式中，管理节点向存储节点发送加速状态命令之后，管理节点监控存储节点的访问带宽。当访问带宽小于阈值或者到达峰值带宽持续时间时，管理节点向存储节点发送退出加速状态命令，退出加速状态命令用于在存储节点将峰值带宽恢复为正常带宽。
本申请实施例中管理节实时监控存储节点的访问带宽，同时存储节点存在多种触发条件退出加速状态，包括突发访问带宽小于阈值或者到达峰值带宽持续时间，使得存储节点能够及时退出加速状态，从而避免了峰值带宽资源的浪费，同时避免损伤存储节点的使用寿命。
一种可能的实施方式中，管理节点根据业务需求信息确定目标配置信息的过程中，管 理节点根据业务需求信息从配置表中确定目标峰值带宽和目标峰值带宽持续时间，配置表包括存储节点的不同配置对应的峰值带宽、峰值带宽持续时间或准备时间，管理节点根据目标峰值带宽和目标峰值带宽持续时间确定目标配置信息，即根据目标峰值带宽和目标峰值带宽持续时间选择响应的GC水线档位、功耗档位和电压档位。
本申请实施例中管理节点能够基于业务需求信息直接从配置表中选择对应的峰值带宽和峰值带宽持续时间，提升了任务配置的效率。
本申请实施例第二方面提供了一种存储节点的配置方法，该方法可以由存储节点执行，也可以由存储节点的部件，例如存储节点的处理器、芯片或芯片系统等执行，还可以由能实现全部或部分存储节点功能的逻辑模块或软件实现。以存储节点执行为例，本申请实施例提供的配置方法包括：存储节点接收管理节点发送的目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。存储节点根据目标配置信息配置存储节点，以使得存储节点处理计算节点对存储节点的突发访问。
本申请实施例中管理节点能够获取计算节点存储节点的访问业务需求，并根据访问业务需求确定存储节点的配置信息，使得存储节点能够在突发访问到来时刻吸收突发访问带宽，提升了存储节点应对突发访问的能力，同时，存储节点没有增加存储介质吸收突发访问带宽，节约了存储资源。
一种可能的实施方式中，目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位或电压档位。配置存储节点的GC水线档位能够使得存储节点在突发访问时刻之前进行垃圾回收，从而避免在突发访问时间到来之后垃圾回收占用访问带宽。配置存储节点的功耗档位能够使得存储节点短时突破功耗限制，从而短时提升存储节点的峰值带宽以吸收突发访问带宽。配置存储节点的电压档位例如配置存储节点的写入电压档位，提升写入电压从而提升计算节点在存储节点的写带宽。
本申请实施例中能够通过多种配置方式配置存储节点的峰值带宽和峰值带宽持续时间，从而提升了配置方案的可实现性。
一种可能的实施方式中，存储节点接收管理节点发送的查询指令，查询指令用于查询存储节点是否到准备完成时刻，准备完成时刻基于突发访问时刻和准备时间确定，准备时间为存储节点的访问带宽到达峰值带宽的所需要的时间。若存储节点到到达准备完成时刻，存储节点向管理节点发送查询指令响应，查询指令响应指示存储节点目标配置信息对应的准备时间已准备完成。
本申请实施例中会接收管理节点发送的查询指令，并向管理节点发送查询指令响应，查询指令响应能够指示存储节点是否准备完成，从提升了存储节点加速到达峰值带宽的可靠性。
一种可能的实施方式中，存储节点接收管理节点发送的加速状态命令，加速状态命令用于在存储节点启动峰值带宽，即加速状态命令能够让目标配置信息对应的峰值带宽和峰值带宽持续时间生效。
本申请实施例中存储节点接收到管理节点发送的加速状态命令之后才将访问带宽提升 至峰值带宽，避免了峰值带宽资源的浪费。
一种可能的实施方式中，存储节点根据目标配置信息配置的过程中，存储节点根据目标配置信息配置存储节点的GC水线档位，增加存储节点在突发访问时刻之前用于进行垃圾处理的带宽。
本申请实施例中配置存储节点的GC水线档位能够使得存储节点在突发访问时刻之前进行垃圾回收，从而避免在突发访问时间到来之后垃圾回收占用访问带宽，从而在突发访问到来时刻增加了用于处理突发访问的带宽资源。
一种可能的实施方式中，存储节点根据目标配置信息配置的过程中，存储节点根据目标配置信息配置存储节点的功耗档位，以使得存储节点功耗阈值增加，从而增加存储节点的带宽。
本申请实施例中配置存储节点的功耗档位能够使得存储节点短时突破功耗限制，从而短时提升存储节点的峰值带宽以吸收突发访问带宽，从而提升了存储节点应对突发访问的能力。
一种可能的实施方式中，存储节点根据目标配置信息配置的过程中，存储节点根据目标配置信息配置存储节点的电压档位，以使得存储节点电压阈值增加，从而增加存储节点的带宽。
本申请实施例中配置存储节点的电压档位，增加存储节点的访问带宽，从而提升了存储节点应对突发访问的能力。
本申请实施例中第三方面提供了一种管理节点，管理节点包括收发单元和处理单元。收发单元用于获取业务需求信息，业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间。处理单元，用于根据业务需求信息确定目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。收发单元还用于向存储节点发送目标配置信息。
一种可能的实施方式中，目标配置信息以下包括一项或多项信息：垃圾回收GC水线档位、功耗档位、电压档位。
一种可能的实施方式中，收发单元具体用于接收计算节点发送的业务需求信息。或者，处理单元还用于根据计算节点对存储节点的历史访问信息确定业务需求信息。
一种可能的实施方式中，当到达突发访问时刻，收发单元还用于向存储节点发送查询指令，查询指令用于查询存储节点是否到准备完成时刻，准备完成时刻基于突发访问时刻和准备时间确定，准备时间为存储节点的访问带宽到达峰值带宽的所需要的时间。当存储节点到达准备完成时刻，则收发单元还用于向存储节点发送加速状态命令，加速状态命令用于在存储节点启动峰值带宽。
一种可能的实施方式中，处理单元还用于监控存储节点的访问带宽。当访问带宽小于阈值或者到达峰值带宽持续时间时，收发单元还用于向存储节点发送退出加速状态命令，退出加速状态命令用于在存储节点将峰值带宽恢复为正常带宽。
一种可能的实施方式中，处理单元具体用于根据业务需求信息从配置表中确定目标峰值带宽和目标峰值带宽持续时间，配置表包括存储节点的不同配置对应的峰值带宽、峰值 带宽持续时间或准备时间。处理单元具体用于根据目标峰值带宽和目标峰值带宽持续时间确定目标配置信息。
本申请实施例第四方面提供了一种存储节点，该存储节点包括收发单元和处理单元。收发单元用于接收管理节点发送的目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。处理单元用于根据目标配置信息配置存储节点，以使得存储节点处理计算节点对存储节点的突发访问。
一种可能的实施方式中，目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位或电压档位。
一种可能的实施方式中，收发单元还用于接收管理节点发送的加速状态命令，加速状态命令用于在存储节点启动峰值带宽。
一种可能的实施方式中，处理单元具体用于根据目标配置信息配置存储节点中SSD的GC水线档位，增加存储节点在突发访问时刻之前用于进行垃圾处理的带宽。
一种可能的实施方式中，处理单元具体用于根据目标配置信息配置存储节点中SSD的功耗档位，以使得存储节点功耗阈值增加，从而增加存储节点的带宽。
一种可能的实施方式中，处理单元具体用于根据目标配置信息配置存储节点中SSD的电压档位，以使得存储节点电压阈值增加，从而增加存储节点的带宽。
本申请实施例第五方面提供了一种存储节点的配置系统，该配置系统包括管理节点和存储节点。其中管理节点用于执行上述第一方面或第一方面任意一种可能的实施方式所述的方法，存储节点用于执行上述第二方面或第二方面任意一种可能的实施方式所述的方法。
本申请实施例第五方面提供了一种配置设备，包括处理器，处理器与存储器耦合，处理器用于存储指令，当指令被处理器执行时，以使得配置设备执行上述第一方面或第一方面任意一种可能的实施方式所述的方法，或者，以使得配置设备执行上述第二方面或第二方面任意一种可能的实施方式所述的方法。
本申请实施例第六方面提供了一种计算机可读存储介质，其上存储有指令，指令被执行时，以使得计算机执行上述第一方面或第一方面任意一种可能的实施方式所述的方法，或者，以使得计算机执行上述第二方面或第二方面任意一种可能的实施方式所述的方法。
本申请实施例第六方面提供了一种计算机程序产品，计算机程序产品中包括指令，其特征在于，指令被执行时，以使得计算机实现上述第一方面或第一方面任意一种可能的实施方式所述的方法，或者，以使得计算机实现上述第二方面或第二方面任意一种可能的实施方式所述的方法。
可以理解，上述提供的任一种管理节点、存储节点、配置系统、配置设备、计算机可读介质或计算机程序产品等所能达到的有益效果可参考对应的方法中的有益效果，此处不再赘述。
附图说明
图1为本申请实施例提供的一种存储系统的系统架构示意图；
图2为本申请实施例提供的一种存储节点的配置方法示意图；
图3为本申请实施例提供的一种存储节点的配置方法流程示意图；
图4为本申请实施例提供的另一种存储节点的配置方法示意图；
图5为本申请实施例提供的一种存储节点的峰值带宽变化示意图；
图6为本申请实施例提供的一种配置装置的结构示意图；
图7为本申请实施例提供的一种配置设备的结构示意图；
图8为本申请实施例提供的一种配置系统的结构示意图。
具体实施方式
本申请实施例提供了一种任务调度方法以及任务调度装置，用于提升任务调度的效率。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”、“第三”、“第四”等(如果存在)是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的实施例能够以除了在这里图示或描述的内容以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
在本申请实施例中，“示例性的”或者“例如”等词用于表示作例子、例证或说明。本申请实施例中被描述为“示例性的”或者“例如”的任何实施例或设计方案不应被解释为比其它实施例或设计方案更优选或更具优势。确切而言，使用“示例性的”或者“例如”等词旨在以具体方式呈现相关概念。
以下，对本申请中的部分用语进行解释说明，以便于本领域技术人员理解。
垃圾回收(garbage collection，GC)是指对于已经分配的内存空间，当程序不再需要使用该内存空间时，就需要销毁对象并释放其所占用的内存资源，使得可以重新利用这段内存空间。
固态硬盘(solid state disk，SSD)是指用固态电子存储芯片阵列制成的硬盘。固态硬盘的存储介质分为两种，一种是采用闪存(flash)作为存储介质，另外一种是采用动态随机存取存储器(dynamic random access memory，DRAM)作为存储介质。
小型计算机系统接口(small computer system interface，SCSI)是一种用于计算机及其周边设备之间系统级接口标准，周边设备包括硬盘、软驱、光驱、打印机或扫描仪等。使用SCSI的硬盘也称作SCSI硬盘。
因特网小型计算机系统接口(internet small computer system interface，iSCSI)是一种基于因特网及SCSI协议的存储技术，把原来只用于本机的SCSI协议透过网络发送，使连接距离可作无限的地域延伸。
以上介绍了本申请实施例中的一些术语，下面结合附图介绍本申请实施例提供的存储节点的配置方法以及相关装置。
请参阅图1，图1为本申请实施例提供的一种存储系统的系统架构示意图。如图1所 示，该存储系统100包括存储节点101、管理节点102和计算节点103。存储节点101包括固态硬盘SSD配置模块1011以及一个或多个SSD1012，其中，SSD配置模块1011用于根据管理节点102发送的配置信息，配置一个或多个SSD1012的峰值带宽和峰值持续时间，配置的具体内容包括垃圾回收GC水线档位、功耗档位或电压档位。
在存储系统100中，多个SSD1012通过分布式软件虚拟为存储池，存储池可以分出不同容量的虚拟云盘，提供给计算节点103使用。具体的，计算节点103上面部署有租户的应用，租户的应用会使用虚拟云盘作为数据盘或系统盘。
管理节点102用于管理存储节点101和计算节点103中的进程，提供告警、监控、日志或配置等功能。管理节点102包括SSD加速配置模块1021，SSD加速配置模块1021用于根据业务需求信息确定存储节点101的配置信息。管理节点102通过小型计算机系统接口SCSI或因特网小型计算机系统接口ISCSI与存储节点101进行通信。
本申请实施例中存储节点101、管理节点102和计算节点103可以分别部署在不同的物理设备上，其中管理节点102也可以部署在虚拟机上，具体不做限定。
在本申请实施例一种可能的系统架构中，存储节点101和管理节点102也可以集成在同一个设备中，当存储节点101和管理节点102集成在同一个设备中时，该设备可以基于一个集成的硬件模块实现上述存储节点101和管理节点102的功能。
请参阅图2，图2为本申请实施例提供的一种配置方法的流程示意图。如图2所示，本申请实施例提供的存储节点的配置方法包括以下步骤：
201.管理节点获取业务需求信息，业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时刻。
管理节点获取业务需求信息，业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时刻。其中，业务需求信息可以是存储节点在未来一段时间内的多个突发访问时刻、突发访问带宽和突发访问持续时刻。例如，存储节点在次日12时到18时这段时间内存在两次突发访问，两次突发访问分别对应不同的突发访问时刻、突发访问带宽和突发访问持续时刻。
本申请实施例中管理节点可以通过多种方式获取业务需求信息，例如，管理节点可以接收计算节点发送的业务需求信息；或者管理节点通过对计算节点进行数据挖掘获得计算节点的业务需求信息；或者管理节点根据计算节点对存储节点的历史访问信息获取业务需求信息的预测值，具体地，管理节点根据历史突发访问时刻、历史突发访问带宽和历史突发访问持续时间预测突发访问时刻、突发访问带宽和突发访问持续时间。
可以理解的是，本申请实施例中的突发访问包括突发写和突发读，若无特殊说明，本申请实施例中的突发访问以突发写为例进行介绍。
请参阅图3，图3为本申请实施例提供的一种存储节点的配置方法流程示意图。在图3所示的步骤1至2中，管理节点根据计算节点的历史访问信息预测得到业务需求信息，例如，管理节点在中午12时判断次日0时至12时是否会发生突发访问，具体的，管理节点根据最近一周的0时至12时计算节点对存储节点的访问情况，预测得到次日0时至12时这段时间内的突发访问时刻、突发访问带宽W以及突发访问持续时间T。
202.管理节点根据业务需求信息确定目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。
管理节点根据业务需求信息确定目标配置信息，目标配置信息包括下述的一种或多种：垃圾回收GC水线档位、功耗档位、电压档位。具体的，管理节点根据业务需求信息从配置表中确定业务需求信息所对应的目标峰值带宽和目标峰值带宽持续时间，选择该目标峰值带宽和目标峰值带宽持续时间所对应的垃圾回收GC水线档位、功耗档位或电压档位作为目标配置信息。
需要说明的是，本申请实施例中存储节点包括一个或多个SSD，管理节点向存储节点发送的目标配置信息具体用于指示存储节点中一个或多个SSD垃圾回收GC水线档位、功耗档位和电压档位。
请参阅表1，表1为本申请实施例提供的一种配置表。如表1所示，在表1的一个示例中，配置表中包括GC水线档位、功耗档位、写电压档位、峰值写带宽、峰值写带宽持续时间或准备时间。从表1中可知，存储节点的峰值写带宽越高，存储节点所需要的准备时间越长，存储节点的峰值写带宽持续时间越长，存储节点所需要的准备时间也越长。存储节点在准备时间内提前垃圾回收、调整存储节点工作模式或进行存储节点内部的数据搬移。
如表1所示，存储节点的GC水线档位包括0、1、2三挡，功耗档位和写入电压档位都包括0、1两档，其中GC水线档位越高，存储节点提前垃圾回收所释放的访问带宽越高。功耗档位越高，存储节点的功耗阈值越高，例如，功耗档位为0时，存储节点的功耗阈值为25W，功耗档位为1时，存储节点的功耗阈值为40W。写电压档位越高，存储节点的写入电压阈值越高。
本申请实施例中配置表可以通过用户手册的方式提供给计算节点，从而方便用户或云厂商查询准备时间。
表1
SSD状态 GC水线档位 功耗档位 写电压档位 峰值写带宽 持续时间 准备时间
状态1 0 0 0 800MB/s 0 0
状态2 1 0 0 4500MB/s 90s 3600s
状态3 2 0 0 4500MB/s 180s 7200s
状态4 0 1 0 1000MB/s 0 60s
状态5 1 1 0 6000MB/s 70s 3600s
状态6 2 1 0 6000MB/s 140s 7200s
状态7 0 0 1 1200MB/s 0 3600s
状态8 1 0 1 6700MB/s 90s 7200s
状态9 2 0 1 6700MB/s 180s 10800s
状态10 0 1 1 1500MB/s 0 3660s
状态11 1 1 1 9000MB/s 70s 7200s
状态12 2 1 1 9000MB/s 140s 10800s
在一个突发写入的示例中，管理节点根据业务需求信息中的突发写带宽和突发写带宽持续时间在配置表中确定出合适的峰值写带宽和峰值写带宽持续时间，并将该峰值写带宽和峰值写带宽持续时间对应的垃圾回收GC水线档位、功耗档位或写电压档位作为目标配置信息。
下面介绍本申请实施例中管理节点在配置表中确定目标配置信息的具体过程。
请参阅图4，图4为本申请实施例提供的确定目标配置信息的示意图。如图4所示，管理节点获取到业务需求信息之后在读取配置表的中峰值带宽和峰值带宽持续时间，假设业务需求信息突发访问带宽为W，突发访问持续时间为T，从配置表中读取的峰值带宽为WS，峰值带宽持续时间为TS。
如图4所示，若配置表中所有WS和TS都满足WS≥W且TS≥T，则所有满足条件的WS和TS中选择WS和TS最小配置作为目标配置信息。
如图4所示，若配置表中存在WS和TS满足WS≥W且TS＜T，则在所有满足条件的配置中选择WS最小的配置，若存在TS使得WS*TS≥W*T，则选择满足条件的配置中选择TS最小的配置，若不存在TS使得WS*TS≥W*T，则选择TS最大的配置。
如图4所示，若配置表中存在WS和TS满足WS＜W且TS≥T，则在所有满足条件的配置中选择WS最大的配置，若存在TS使得WS*TS≥W*T，则选择满足条件的配置中TS最小的配置，若不存在TS使得WS*TS≥W*T，则TS选择最大的配置。
如图4所示，若配置表中的所有WS和TS满足WS＜W且TS＜T，则选择WS和TS都最大的配置作为目标配置信息。
在图4所示的一个示例中，假设突发访问带宽W为5500MB/S，突发访问持续时间T为200S，管理节点从配置表中确定目标配置信息，配置表以表1为例，可以表1中存在WS≥W且TS＜T的配置，其中，满足峰值带宽WS≥W的配置包括状态5、状态6、状态8、状态9、状态11和状态12。在这些配置中，满足WS*TS≥W*T的配置包括状态9和状态12，其中状态9的WS＜状态12的WS，因此选择状态9对应的配置作为目标配置信息。
请继续参阅图3，在图3所示的步骤3中，管理节点根据业务需求信息确定目标配置信息。例如，业务需求信息指示突发访问时刻为中午12点，突发写带宽为5000MB/s，持续时间为60s。管理节点在配置表中查询峰值写带宽和峰值写带宽持续时间，配置表以表1为例，SSD的所有状态中，峰值写带宽大于或等于5000MB/s且峰值写带宽持续时间大于或等于60s的选项有状态5、状态6、状态8、状态9、状态11和状态12这六种，这六种选项中数据量最小的是状态5，数据量为峰值写带宽和峰值写带宽持续时间的乘积，因此，管理节点将状态5对应的GC水线档位、功耗档位和电压档位作为目标配置信息。
203.管理节点向存储节点发送目标配置信息。
管理节点向存储节点发送目标配置信息，具体的，管理节点通过SSD峰值性能配置接口向存储节点发送的目标配置信息中的一下一项或多项信息：GC水线档位、功耗档位、电压档位。
本申请实施中存储节点接收到管理节点发送的目标配置信息之后就可以进入准备状态，或者，存储节点接收到管理节点的发送的准备指令之后再进入准备状态，具体不做限 定。准备状态下的存储节点开始准备时间计时。
请继续参阅图3，在图3所示的步骤4中，管理节点向存储节点发送目标配置信息。例如，目标配置信息为表1中状态5对应的GC水线档位、功耗档位和电压档位，管理节点将状态5对应的配置发送到存储节点。存储节点接收到表1中状态5对应的配置之后，由于进入状态5需要的准备时间为3600s，因此，存储节点需要至少提前3600s进入状态5的准备状态。
在图3所示的示例中，以存储节点接收到管理节点的发送的准备指令之后再进入准备状态为例，假设突发访问时刻为12:00，在10:30时，管理节点向存储节点发送进入状态5的准备指令。
本申请实施例中提供了一种峰值带宽可配置的存储节点，同时存储节点提供了SSD峰值性能配置接口，从而使得管理节点根据目标配置信息调用SSD峰值性能配置接口改变存储的峰值写带宽。
204.存储节点根据目标配置信息配置存储节点。
存储节点根据目标配置信息配置存储节点，包括配置存储节点的GC水线档位、功耗档位和电压档位，上述几种存储节点的配置方法可以分时配置，也可以同时叠加配置，具体不做限定。下面分别具体介绍上述几种配置方法。
存储节点根据目标配置信息配置存储节点的GC水线档位的过程中，存储节点在准备时间提前开始垃圾回收，从而避免在突发访问时间到来之后垃圾回收占用访问带宽。具体的，存储节点预留一部分空白块的空间用于吸收突发负载，吸收突发负载的数据量由GC水线档位确定。空白块是指未存储数据的存储空间，块为数据存储和处理的单位。当突发负载来临时，空白块能够以接近正常写带宽的速度吸收突发写带宽，直到空白块的空间消耗完成，顺序写带宽例如4.5GB/S。
存储节点根据目标配置信息配置存储节点的功耗档位的过程中，存储节点根据目标配置信息提高存储节点的功耗档位，使得存储节点短时突破功耗限制，从而短时提升存储节点的峰值带宽吸收突发访问带宽。例如，存储节点的SSD在常规状态的下功耗阈值为25W，存储节点根据目标配置信息将SSD的功耗阈值提升至40W，存储节点的SSD写带宽被充分释放，存储节点的些带宽能够从4.5GB/S提升至6GB/S。
存储节点根据目标配置信息配置存储节点的电压档位的过程中，以写电压为例，存储节点可以调节SSD的闪存颗粒的写电压阈值，写电压阈值越高，写电压步幅越高，从而存储节点的SSD到达指定状态的步数越少，写入带宽越高。例如，电压档位为0档代表正常写电压档，电压档位为1档时为高写电压档，高写电压档相较于正常写电压档，写带宽能够提升50％。
本申请实施例中存储节点根据目标配置信息配置完成存储节点之后，当计算节点对存储节点的突发访问到来时，管理节点向存储节点发送查询指令，查询指令用于查询存储节点是否到达准备完成时刻，即查询存储节点的准备时间计时是否计时完成，若准备计时完成，则管理节点确定存储节点准备完成。具体的，存储节点可以向管理节点发送查询指令响应，该查询指令响应用于向管理节点指示存储节点的准备状态。
管理节点确定存储节点到达准备完成时刻之后，存储节点向管理节点发送加速状态命令，该加速状态命令用于设置存储节点开启加速状态，开始加速之后，存储节点的访问带宽到达峰值带宽。管理节点实时监控存储节点的访问带宽，当存储节点的突发访问带宽结束，或者，峰值带宽的持续时间已过，则管理节点向存储节点发送退出加速状态命令，存储节点接收退出加速状态命令之后，存储节点的带宽从峰值带宽恢复到正常带宽。
请继续参阅图3，在图3所示的步骤5至10中，存储节点接收到管理节点发送的目标配置信息之后，存储节点进入准备状态。当到达突发访问时刻，管理节点向存储节点发送查询指令，查询存储节点是否准备完成。若存储节点准备完成，则管理节点向存储节点发送加速状态命令。若未准备完成，则存储节点未按照目标配置信息配置成功，存储节点不具备峰值调节能力。
在图3所示的示例中，管理节点持续监控存储节点的访问带宽，若突发访问结束或者峰值带宽的持续时间已过，则管理节点向存储节点发送退出加速状态命令，使得存储节点由峰值带宽恢复至正常带宽。
请参阅图5，图5为本申请实施例提供的一种存储节点的峰值写带宽的变化示意图。如图5所示，存储节点的峰值写带宽根据目标配置信息配置成功之后，当存储节点接收到加速状态命令之后，存储节点的峰值带宽会提高，当存储节点接收到退出加速状态命令之后，存储节点的峰值带宽会降低到峰值带宽的正常值。
如图5所示，存储节点的SSD的常规峰值写带宽值为W0，在Ts时刻发送目标配置信息，根据目标配置信息将SSD的峰值写带宽配置成为W1，并且持续T1时间。W1和T1对应的准备时间为(T0-Ts)，经过(T0-Ts)准备时间之后，在T0时刻，SSD的峰值写带宽达到W1，并且可持续时间为T1。在Ts’时刻，SSD的峰值写带宽被配置成为W2，并且持续时间为T2，W2和T2对应的准备时间为(T0’-Ts’)，经过(T0’-Ts’)准备时间之后，SSD的峰值写带宽达到W2，并且持续时间为T2。
本申请实施例中管理节点能够获取计算节点的访问业务需求，并根据访问业务需求确定存储节点的配置信息，使得存储节点能够在突发访问到来时刻吸收突发访问带宽，提升了存储节点应对突发访问的能力，同时，存储节点没有增加新的存储介质吸收突发访问带宽，降低存储节点应对突发访问的成本，节约了存储资源。
上面介绍本申请实施例提供的存储节点的配置方法，下面结合附图介绍本申请实施例涉及的配置装置。
请参阅图6，图6为本申请实施例提供的一种配置装置的结构示意图。该装置用于实现上述各实施例中管理节点或者存储节点所执行的各个步骤，如图6所示，该配置装置600包括收发单元601和处理单元602。
在本申请的一个实施例中，配置装置600用于实现上述方法实施例中管理节点所执行的方法，即管理节点可以包括收发单元601和处理单元602。
其中，收发单元601用于获取业务需求信息，业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间。处理单元602用于根据业务需求信息确定目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。 收发单元还用于向存储节点发送目标配置信息。
一种可能的实施方式中，目标配置信息以下包括一项或多项信息：垃圾回收GC水线档位、功耗档位、电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
一种可能的实施方式中，收发单元601具体用于接收计算节点发送的业务需求信息。或者，处理单元还用于根据计算节点对存储节点的历史访问信息确定业务需求信息的预测值。
一种可能的实施方式中，当到达突发访问时刻，收发单元601还用于向存储节点发送查询指令，查询指令用于查询存储节点是否到准备完成时刻，准备完成时刻基于突发访问时刻和准备时间确定，准备时间为存储节点的访问带宽到达峰值带宽的所需要的时间。当存储节点到达准备完成时刻，则收发单元601还用于向存储节点发送加速状态命令，加速状态命令用于在存储节点启动峰值带宽。
一种可能的实施方式中，处理单元602还用于监控存储节点的访问带宽。当访问带宽小于阈值或者到达峰值带宽持续时间时，收发单元601还用于向存储节点发送退出加速状态命令，退出加速状态命令用于在存储节点将峰值带宽恢复为正常带宽。
一种可能的实施方式中，处理单元602具体用于根据业务需求信息从配置表中确定目标峰值带宽和目标峰值带宽持续时间，配置表包括存储节点的不同配置对应的峰值带宽、峰值带宽持续时间或准备时间。处理单元602具体用于根据目标峰值带宽和目标峰值带宽持续时间确定目标配置信息。
在本申请的另一个实施例中，配置装置600用于实现上述方法实施例中存储节点所执行的方法，即存储节点可以包括收发单元601和处理单元602。
其中收发单元601用于接收管理节点发送的目标配置信息，目标配置信息用于配置存储节点的峰值带宽和峰值带宽持续时间。处理单元602用于根据目标配置信息配置存储节点，以使得存储节点处理计算节点对存储节点的突发访问。
一种可能的实施方式中，目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位或电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
一种可能的实施方式中，收发单元601还用于接收管理节点发送的加速状态命令，加速状态命令用于在存储节点启动峰值带宽。
一种可能的实施方式中，处理单元602具体用于根据目标配置信息配置存储节点中SSD的GC水线档位，增加存储节点在突发访问时刻之前用于进行垃圾处理的带宽。
一种可能的实施方式中，处理单元602具体用于根据目标配置信息配置存储节点中SSD的功耗档位，以使得存储节点功耗阈值增加，从而增加存储节点的带宽。
一种可能的实施方式中，处理单元602具体用于根据目标配置信息配置存储节点中SSD的电压档位，以使得存储节点电压阈值增加，从而增加存储节点的带宽。
应理解以上装置中单元的划分仅仅是一种逻辑功能的划分，实际实现时可以全部或部分集成到一个物理实体上，也可以物理上分开。且装置中的单元可以全部以软件通过处理元件调用的形式实现；也可以全部以硬件的形式实现；还可以部分单元以软件通过处理元件调用的形式实现，部分单元以硬件的形式实现。例如，各个单元可以为单独设立的处理 元件，也可以集成在装置的某一个芯片中实现，此外，也可以以程序的形式存储于存储器中，由装置的某一个处理元件调用并执行该单元的功能。此外这些单元全部或部分可以集成在一起，也可以独立实现。这里所述的处理元件又可以成为处理器，可以是一种具有信号的处理能力的集成电路。在实现过程中，上述方法的各步骤或以上各个单元可以通过处理器元件中的硬件的集成逻辑电路实现或者以软件通过处理元件调用的形式实现。
值得说明的是，对于上述方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员应该知悉，本发明本申请并不受所描述的动作顺序的限制，其次，本领域技术人员也应该知悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作并不一定是本发明本申请所必须的。
本领域的技术人员根据以上描述的内容，能够想到的其他合理的步骤组合，也属于本发明本申请的保护范围内。其次，本领域技术人员也应该熟悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作并不一定是本发明本申请所必须的。
请参阅图7，图7为本申请实施例提供的一种配置设备示意图。如图7所示，该配置设备700包括：处理器710、存储器720和接口730，处理器710、存储器720与接口730通过总线(图中未标注)耦合。存储器720存储有指令，当存储器720中的执行指令被执行时，配置设备700执行上述方法实施例中管理节点或存储接节点所执行的方法。
配置设备700可以是被配置成实施以上方法的一个或多个集成电路，例如：一个或多个特定集成电路(application specific integrated circuit，ASIC)，或，一个或多个微处理器(digital singnal processor，DSP)，或，一个或者多个现场可编程门阵列(field programmable gate array，FPGA)，或这些集成电路形式中至少两种的组合。再如，当装置中的单元可以通过处理元件调度程序的形式实现时，该处理元件可以是通用处理器，例如中央处理器(central processing unit，CPU)或其它可以调用程序的处理器。再如，这些单元可以集成在一起，以片上系统(system-on-a-chip，SOC)的形式实现。
处理器710可以是中央处理单元(central processing unit，CPU)，还可以是其它通用处理器、数字信号处理器(digital signal processor，DSP)、专用集成电路(application specific integrated circuit，ASIC)、现场可编程门阵列(field programmable gate array，FPGA)或者其它可编程逻辑器件、晶体管逻辑器件，硬件部件或者其任意组合。通用处理器可以是微处理器，也可以是任何常规的处理器。
存储器720可以包括只读存储器和随机存取存储器，并向处理器710提供指令和数据。存储器720还可以包括非易失性随机存取存储器。例如，存储器720可设置多个分区，每个区域分别用于存储不同软件模块的私钥。
存储器720可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(read-only memory，ROM)、可编程只读存储器(programmable ROM，PROM)、可擦除可编程只读存储器(erasable PROM，EPROM)、电可擦除可编程只读存储器(electrically EPROM，EEPROM)或闪存。易失性存储器可以是随机存取存储器(random access memory，RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(static RAM， SRAM)、动态随机存取存储器(DRAM)、同步动态随机存取存储器(synchronous DRAM，SDRAM)、双倍数据速率同步动态随机存取存储器(double data date SDRAM，DDR SDRAM)、增强型同步动态随机存取存储器(enhanced SDRAM，ESDRAM)、同步连接动态随机存取存储器(synchlink DRAM，SLDRAM)和直接内存总线随机存取存储器(direct rambus RAM，DR RAM)。
总线除包括数据总线之外，还可以包括电源总线、控制总线和状态信号总线等。总线可以是快捷外围部件互连标准(peripheral component interconnect express，PCIe)总线，或扩展工业标准结构(extended industry standard architecture，EISA)总线、统一总线(unified bus，Ubus或UB)、计算机快速链接(compute express link，CXL)、缓存一致互联协议(cache coherent interconnect for accelerators，CCIX)等。总线可以分为地址总线、数据总线、控制总线等。
请参阅图8，图8是本申请实施例提供的一种配置系统示意图，该配置系统800包括管理节点801和存储节点802，管理节点801可以是上述方法实施例中的管理节点，存储节点802可以是上述方法实施例中的存储节点。
在本申请的另一个实施例中，还提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，当设备的处理器执行该计算机执行指令时，设备执行上述方法实施例中管理节点或存储接节点所执行的方法。
在本申请的另一个实施例中，还提供一种计算机程序产品，该计算机程序产品包括计算机执行指令，该计算机执行指令存储在计算机可读存储介质中。当设备的处理器执行该计算机执行指令时，设备执行上述方法实施例中管理节点或存储接节点所执行的方法。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统，装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统，装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出 来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，read-only memory)、随机存取存储器(RAM，random access memory)、磁碟或者光盘等各种可以存储程序代码的介质。

Claims (28)

  1. 一种存储节点的配置方法，其特征在于，包括：
    获取业务需求信息，所述业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间；
    根据所述业务需求信息确定目标配置信息，所述目标配置信息用于配置所述存储节点的峰值带宽和峰值带宽持续时间；
    向所述存储节点发送所述目标配置信息。
  2. 根据权利要求1所述的方法，其特征在于，所述目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位、电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
  3. 根据权利要求1或2所述的方法，其特征在于，所述获取业务需求信息包括：
    接收所述计算节点发送的所述业务需求信息；
    或者，根据所述计算节点对所述存储节点的历史访问信息确定所述业务需求信息的预测值。
  4. 根据权利要求1至3中任一项所述的方法，其特征在于，所述向所述存储节点发送所述目标配置信息之后，所述方法还包括：
    当到达所述突发访问时刻，向所述存储节点发送查询指令，所述查询指令用于查询所述存储节点是否到准备完成时刻，所述准备完成时刻基于所述突发访问时刻和准备时间确定，所述准备时间为所述存储节点的访问带宽到达所述峰值带宽的所需要的时间；
    当所述存储节点到达准备完成时刻，则向所述存储节点发送加速状态命令，所述加速状态命令用于在所述存储节点启动所述峰值带宽。
  5. 根据权利要求4所述的方法，其特征在于，所述向所述存储节点发送加速状态命令之后，所述方法还包括：
    监控所述存储节点的访问带宽；
    当所述访问带宽小于阈值或者到达峰值带宽持续时间时，向所述存储节点发送退出加速状态命令，所述退出加速状态命令用于在所述存储节点将所述峰值带宽恢复为正常带宽。
  6. 根据权利要求1至5中任一项所述的方法，其特征在于，所述根据所述业务需求信息确定目标配置信息包括：
    根据所述业务需求信息从配置表中确定目标峰值带宽和目标峰值带宽持续时间，所述配置表包括所述存储节点的不同配置对应的峰值带宽、峰值带宽持续时间或准备时间；
    根据所述目标峰值带宽和目标峰值带宽持续时间确定所述目标配置信息。
  7. 一种存储节点的配置方法，其特征在于，包括：
    接收目标配置信息，所述目标配置信息用于配置所述存储节点的峰值带宽和峰值带宽持续时间；
    根据所述目标配置信息配置存储节点，以使得所述存储节点处理所述计算节点对所述存储节点的突发访问。
  8. 根据权利要求7所述的方法，其特征在于，所述目标配置信息包括以下一项或多项 信息：垃圾回收GC水线档位、功耗档位或电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
  9. 根据权利要求7或8所述的方法，其特征在于，所述方法还包括：
    接收加速状态命令，所述加速状态命令用于在所述存储节点启动所述峰值带宽。
  10. 根据权利要求7至9中任一项所述的方法，其特征在于，所述根据所述目标配置信息配置所述存储节点包括：
    根据所述目标配置信息配置所述存储节点中固态硬盘SSD的GC水线档位，增加所述存储节点在突发访问时刻之前用于进行垃圾处理的带宽。
  11. 根据权利要求7至10中任一项所述的方法，其特征在于，所述根据所述目标配置信息配置所述存储节点包括：
    根据所述目标配置信息配置所述存储节点中SSD的功耗档位，以使得所述存储节点功耗阈值增加，从而增加所述存储节点的带宽。
  12. 根据权利要求7至11任一项所述的方法，其特征在于，所述根据所述目标配置信息配置所述存储节点包括：
    根据所述目标配置信息配置所述存储节点中SSD的电压档位，以使得所述存储节点电压阈值增加，从而增加所述存储节点的带宽。
  13. 一种管理节点，其特征在于，包括：
    收发单元，用于获取业务需求信息，所述业务需求信息包括计算节点对存储节点的突发访问时刻、突发访问带宽或突发访问持续时间；
    处理单元，用于根据所述业务需求信息确定目标配置信息，所述目标配置信息用于配置所述存储节点的峰值带宽和峰值带宽持续时间；
    所述收发单元还用于向所述存储节点发送所述目标配置信息。
  14. 根据权利要求13所述的管理节点，其特征在于，所述目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位、电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
  15. 根据权利要求13或14所述的管理节点，其特征在于，所述收发单元具体用于接收所述计算节点发送的所述业务需求信息；
    或者，所述处理单元还用于根据所述计算节点对所述存储节点的历史访问信息确定所述业务需求信息的预测值。
  16. 根据权利要求13至15任一项所述的管理节点，其特征在于，当到达所述突发访问时刻，所述收发单元还用于向所述存储节点发送查询指令，所述查询指令用于查询所述存储节点是否到准备完成时刻，所述准备完成时刻基于所述突发访问时刻和准备时间确定，所述准备时间为所述存储节点的访问带宽到达所述峰值带宽的所需要的时间；
    当所述存储节点到达准备完成时刻，则所述收发单元还用于向所述存储节点发送加速状态命令，所述加速状态命令用于在所述存储节点启动所述峰值带宽。
  17. 根据权利要求16所述的管理节点，其特征在于，所述处理单元还用于监控所述存储节点的访问带宽；
    当所述访问带宽小于阈值或者到达峰值带宽持续时间时，所述收发单元还用于向所述存储节点发送退出加速状态命令，所述退出加速状态命令用于在所述存储节点将所述峰值带宽恢复为正常带宽。
  18. 根据权利要求13至17中任一项所述的管理节点，其特征在于，所述处理单元具体用于根据所述业务需求信息从配置表中确定目标峰值带宽和目标峰值带宽持续时间，所述配置表包括所述存储节点的不同配置对应的峰值带宽、峰值带宽持续时间或准备时间；
    所述处理单元具体用于根据所述目标峰值带宽和目标峰值带宽持续时间确定所述目标配置信息。
  19. 一种存储节点，其特征在于，包括：
    收发单元，用于接收管理节点发送的目标配置信息，所述目标配置信息用于配置所述存储节点的峰值带宽和峰值带宽持续时间；
    处理单元，用于根据所述目标配置信息配置所述存储节点，以使得所述存储节点处理所述计算节点对所述存储节点的突发访问。
  20. 根据权利要求19所述的存储节点，其特征在于，所述目标配置信息包括以下一项或多项信息：垃圾回收GC水线档位、功耗档位或电压档位，其中，垃圾回收水线指示进行垃圾回收的数据量。
  21. 根据权利要求19或20所述的存储节点，其特征在于，所述收发单元还用于接收所述管理节点发送的加速状态命令，所述加速状态命令用于在所述存储节点启动所述峰值带宽。
  22. 根据权利要求19至21中任一项所述的存储节点，其特征在于，所述处理单元具体用于根据所述目标配置信息配置所述存储节点中SSD的GC水线档位，增加所述存储节点在突发访问时刻之前用于进行垃圾处理的带宽。
  23. 根据权利要求19至22中任一项所述的存储节点，其特征在于，所述处理单元具体用于根据所述目标配置信息配置所述存储节点中SSD的功耗档位，以使得所述存储节点功耗阈值增加，从而增加所述存储节点的带宽。
  24. 根据权利要求19至23任一项所述的存储节点，其特征在于，所述处理单元具体用于根据所述目标配置信息配置所述存储节点中SSD的电压档位，以使得所述存储节点电压阈值增加，从而增加所述存储节点的带宽。
  25. 一种存储节点的配置系统，包括：管理节点和存储节点；
    所述管理节点用于执行上述权利要求1至6中任意一项所述的方法，所述存储节点用于执行上述权利要求7至12中任一项所述的方法。
  26. 一种配置设备，其特征在于，包括处理器，所述处理器与存储器耦合，所述处理器用于存储指令，当所述指令被所述处理器执行时，以使得所述配置设备执行权利要求1至6中任一项所述的方法，或者，以使得所述配置设备执行权利要求7至12中任一项所述的方法。
  27. 一种计算机可读存储介质，其上存储有指令，其特征在于，所述指令被执行时，以使得计算机执行权利要求1至6中任一项所述的方法，或者，以使得计算机执行权利要求 7至12中任一项所述的方法。
  28. 一种计算机程序产品，所述计算机程序产品中包括指令，其特征在于，所述指令被执行时，以使得计算机实现权利要求1至6中任一项所述的方法，或者，以使得计算机执行权利要求7至12中任一项所述的方法。
PCT/CN2022/123763 2021-12-31 2022-10-08 一种存储节点的配置方法以及相关装置 WO2023124347A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP22913612.2A EP4443833A1 (en) 2021-12-31 2022-10-08 Method for configuring storage node, and related apparatus
US18/759,328 US20240354247A1 (en) 2021-12-31 2024-06-28 Storage Node Configuration Method and Related Apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111679415.2A CN116418680A (zh) 2021-12-31 2021-12-31 一种存储节点的配置方法以及相关装置
CN202111679415.2 2021-12-31

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/759,328 Continuation US20240354247A1 (en) 2021-12-31 2024-06-28 Storage Node Configuration Method and Related Apparatus

Publications (1)

Publication Number Publication Date
WO2023124347A1 true WO2023124347A1 (zh) 2023-07-06

Family

ID=86997449

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/123763 WO2023124347A1 (zh) 2021-12-31 2022-10-08 一种存储节点的配置方法以及相关装置

Country Status (4)

Country Link
US (1) US20240354247A1 (zh)
EP (1) EP4443833A1 (zh)
CN (1) CN116418680A (zh)
WO (1) WO2023124347A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108768877A (zh) * 2018-07-20 2018-11-06 网宿科技股份有限公司 一种突发流量的分配方法、装置及代理服务器
CN111078588A (zh) * 2019-12-17 2020-04-28 北京三快在线科技有限公司 垃圾回收方法、装置、设备及存储介质
CN111131516A (zh) * 2019-12-31 2020-05-08 中国联合网络通信集团有限公司 业务的配置方法及装置
CN111338782A (zh) * 2020-03-06 2020-06-26 中国科学技术大学 面向共享式突发数据缓存的基于竞争感知的节点分配方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108768877A (zh) * 2018-07-20 2018-11-06 网宿科技股份有限公司 一种突发流量的分配方法、装置及代理服务器
CN111078588A (zh) * 2019-12-17 2020-04-28 北京三快在线科技有限公司 垃圾回收方法、装置、设备及存储介质
CN111131516A (zh) * 2019-12-31 2020-05-08 中国联合网络通信集团有限公司 业务的配置方法及装置
CN111338782A (zh) * 2020-03-06 2020-06-26 中国科学技术大学 面向共享式突发数据缓存的基于竞争感知的节点分配方法

Also Published As

Publication number Publication date
EP4443833A1 (en) 2024-10-09
CN116418680A (zh) 2023-07-11
US20240354247A1 (en) 2024-10-24

Similar Documents

Publication Publication Date Title
US9229525B2 (en) Adaptive latency tolerance for power management of memory bus interfaces
US9110669B2 (en) Power management of a storage device including multiple processing cores
US7966506B2 (en) Saving power in a computer system
EP3005079B1 (en) On-the-fly performance adjustment for solid state storage devices
US10860494B2 (en) Flushing pages from solid-state storage device
KR20160114726A (ko) 명령어 큐잉
TWI639921B (zh) 指令處理方法及使用所述方法的儲存控制器
US11175858B2 (en) Memory system control method receiving optimized buffer flush/fill (OBFF) messages over a PCIE bus
US20190041957A1 (en) Technologies for providing efficient detection of idle poll loops
US20160239206A1 (en) Semiconductor system performing status read for semiconductor device and operating method thereof
WO2014206078A1 (zh) 内存访问方法、装置及系统
US8639840B2 (en) Processing unit, chip, computing device and method for accelerating data transmission
WO2023051000A1 (zh) 内存管理方法、装置、处理器及计算设备
US12079139B2 (en) Hierarchical memory systems
US20230385190A1 (en) Communication method, apparatus, and system
CN115495433A (zh) 一种分布式存储系统、数据迁移方法及存储装置
US20240231654A1 (en) Method and Apparatus for Controlling Internal Memory Bandwidth, Processor, and Computing Device
WO2014180443A1 (zh) 系统资源均衡调整方法及装置
US9781225B1 (en) Systems and methods for cache streams
WO2023124347A1 (zh) 一种存储节点的配置方法以及相关装置
CN112445412B (zh) 一种数据存储方法及装置
Jian et al. Understanding and optimizing power consumption in memory networks
US9910774B1 (en) Spontaneous reconfiguration of data structures using balloon memory allocation
CN104750425A (zh) 一种存储系统及其非易失性存储器的控制方法
WO2019223445A1 (zh) 硬盘读写控制方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22913612

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2022913612

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2022913612

Country of ref document: EP

Effective date: 20240705

NENP Non-entry into the national phase

Ref country code: DE