WO2014180418A1 - 负载均衡方法和装置 - Google Patents

负载均衡方法和装置 Download PDF

Info

Publication number
WO2014180418A1
WO2014180418A1 PCT/CN2014/079002 CN2014079002W WO2014180418A1 WO 2014180418 A1 WO2014180418 A1 WO 2014180418A1 CN 2014079002 W CN2014079002 W CN 2014079002W WO 2014180418 A1 WO2014180418 A1 WO 2014180418A1
Authority
WO
WIPO (PCT)
Prior art keywords
load
board
boards
indicator
new user
Prior art date
Application number
PCT/CN2014/079002
Other languages
English (en)
French (fr)
Inventor
张广军
王奎
宋劼明
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2014180418A1 publication Critical patent/WO2014180418A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/64Hybrid switching systems
    • H04L12/6418Hybrid transport

Definitions

  • the present invention relates to the field of communications, and in particular to a load balancing method and apparatus.
  • CDN content delivery network
  • the signaling processing module needs to select a board that provides services for users when the user accesses. For the controllability of the process, this selection is often one-time. Therefore, it is especially important to choose the board that serves the user and ensure the load balancing between the boards to reduce the excessive load on the board.
  • a load balancing scheme of a CDN system in the related art adopts a round-robin strategy, that is, a board that selects services is selected in turn on all boards, and this method can easily achieve equalization on the number of users, and is normally compared. Stable, without the need to provide external data (such as the load data of the board) can be achieved.
  • the inventors found that when using the wheel selection strategy, problems occur in the following situations:
  • the embodiments of the present invention provide a load balancing method and apparatus to address at least the problem of load imbalance caused by a rounding policy or a preferred policy of a content distribution network in the related art.
  • a load balancing method including: determining, according to a load of a plurality of boards of a content distribution network, a load level corresponding to each of the plurality of boards; One of the one or more boards corresponding to the first load level in the load level provides service for the new user, where the load of the one or more boards corresponding to the first load level is Below the load corresponding to other load levels in the load level.
  • the method further includes: determining the load of the plurality of boards separately according to a predetermined load indicator.
  • the predetermined load indicator includes at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board.
  • selecting the one of the one or more boards corresponding to the first load level comprises: determining whether a load indicator of the one or more boards meets a load of the new user requirement The indicator, where the load indicator of the new user requirement includes at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board; A board in which the load indicator meets the load indicator of the new user requirement is provided for the new user.
  • determining the load level corresponding to each of the plurality of boards includes: the plurality of boards according to a plurality of threshold ranges in which the loads of the plurality of boards respectively fall And mapping to a plurality of queues, wherein one of the plurality of queues respectively corresponds to one of the load levels.
  • selecting the one of the one or more boards corresponding to the first load level comprises: sequentially determining a load of one or more boards in the queue corresponding to the first load level Whether the metric meets the load metric of the new user requirement, where the load metric of the new user requirement includes at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processing of the board.
  • the first load board that meets the load indicator of the new user requirement in the queue corresponding to the first load level is used to provide the new user.
  • the method further includes: updating the load level of the corresponding one of the plurality of boards in at least one of the following: reaching a predetermined update period; the corresponding board starts to be new The user provides the service; the user of the service provided by the corresponding board exits; the corresponding board is powered off or an abnormality occurs, wherein the load of the corresponding board that is powered off or abnormal is determined to be full. load.
  • a load balancing apparatus including: a determining module, configured to determine each of the plurality of boards according to a load of multiple boards of the content distribution network Corresponding load level; a selection module, configured to select one of the one or more boards corresponding to the first load level in the load level to provide service for the new user, where the first load level corresponds to The load of the one or more boards is lower than the load corresponding to other load levels in the load level.
  • the selection module includes: a determining unit, configured to determine whether a load indicator of the one or more boards meets a load indicator of the new user requirement, where the load indicator of the new user requirement includes at least the following One of the file system input and output indicators of the board, the network port traffic indicator of the board, and the CPU load indicator of the board; the selection unit is set to select the load indicator of the one or more boards to meet the new one.
  • a board of the load indicator required by the user serves the new user.
  • the device further includes: an update module, configured to update the load level of the corresponding one of the plurality of boards in at least one of the following: reaching a predetermined update period;
  • the board starts to provide services for the new user; the user of the service provided by the corresponding board exits; the corresponding board is powered off or abnormal, and the corresponding single is determined to be powered off or abnormal.
  • the board load is full.
  • the load level corresponding to each of the plurality of boards is determined according to the load of the plurality of boards of the content distribution network; and one or more orders corresponding to the first load level in the load level are selected.
  • a board in the board provides services for new users, where the load of one or more boards corresponding to the first load level is lower than the load corresponding to other load levels in the load level, and the content in the related art is solved.
  • the load balancing problem caused by the rounding policy or the preferred policy of the distribution network is implemented, thereby implementing load balancing of multiple boards of the content distribution network.
  • FIG. 1 is a schematic flow chart of a load balancing method according to an embodiment of the present invention
  • 2 is a schematic structural diagram of a load balancing apparatus according to an embodiment of the present invention
  • FIG. 3 is a schematic diagram of a preferred structure of a load balancing apparatus according to an embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of a load balancing apparatus according to an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of a functional module topology according to a preferred embodiment of the present invention
  • FIG. 6 is a flow chart showing a hierarchical control process according to a preferred embodiment of the present invention
  • FIG. 7 is a schematic diagram of a hierarchical queue according to a preferred embodiment of the present invention. .
  • FIG. 1 is a schematic flowchart of a load balancing method according to an embodiment of the present invention. As shown in FIG.
  • the method includes the following process: Step S102: Multiple orders according to a content distribution network
  • the load of the board determines the load level corresponding to each of the plurality of boards.
  • the load of one or more boards corresponding to the first load level is lower than the load corresponding to the other load levels in the load level.
  • the load level corresponding to each board is determined according to the load of the multiple boards, and the load level is selected as one of the one or more boards in the first load level with low load.
  • Serve new users The solution shown in the above steps is based on a single-board load balancing policy.
  • the solution can achieve load balancing with respect to the round-robin strategy in the related art; since the above solution is based on one or more of the same load levels.
  • the equalization strategy of the board therefore, relative to the preferred strategy in the related art, the requirement of the coefficient and the timeliness and accuracy of the data are relatively low, and the load of the single board with the lightest load before is suddenly increased. problem.
  • the above steps in the related art solve the problem in the related art.
  • Load balancing of multiple boards in the content distribution network is achieved by the problem of load imbalance caused by the distribution policy of the distribution network or the preferred policy.
  • the load of each of the plurality of boards is determined according to a predetermined load indicator.
  • the predetermined load indicator may be an indicator of at least one of the following: a file system input and output indicator of the board, and a network port of the board. Traffic metrics, CPU load metrics for the board.
  • the predetermined load indicator is not limited to the above-mentioned indicators, and the user may set other load indicators according to requirements or requirements of the system or separately select or combine the load indicators.
  • the load indicator that is easy to cause the overall load bottleneck is selected for the board. For example, for a type of board, the processing power of the CPU is relatively strong and the indicator is not easy to appear. If the load is too high, the main factor causing the load is too large. The network port traffic indicator is too high.
  • the network port traffic indicator can be preferably included as the predetermined load indicator.
  • the multiple load indicators of the multiple boards can be normalized, and the normalized processing is performed on different boards or different types of boards in the load balancing processing of the same content distribution network.
  • Different predetermined load indicators are used to determine the load of the respective boards.
  • different boards or different types of boards respectively select a load indicator that easily causes an overall load bottleneck thereof as the predetermined load indicator.
  • the selected board in order to enable a new user to be added to serve a board after selecting a board, the selected board does not have a problem of excessive load, and may be corresponding to one of the first load levels.
  • the load indicator of the new user requirement includes, but is not limited to, at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board.
  • a plurality of boards may be sequentially mapped to a plurality of queues according to a plurality of threshold ranges in which the load of the plurality of boards respectively falls, wherein one of the plurality of queues respectively corresponds to the load level.
  • a load level A load level.
  • the threshold range of the plurality of threshold ranges corresponds to a load level, and the number of the threshold range or the load level may be selected by the user according to the granularity and precision of load balancing required, and the threshold number or the number of load levels is increased.
  • the load balancing capability of the load balancing method is stronger, but the real-time and accuracy requirements of the required data will be larger.
  • step S104 it may be determined in sequence whether the load indicator of one or more boards in the queue corresponding to the first load level meets the load indicator of the new user requirement, and The first board that meets the load indicator of the new user requirement in the queue corresponding to the first load level is selected to serve the new user.
  • the load indicator of the new user includes at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board; preferably, the first load level is determined.
  • one of the second load levels with a higher load than the first load level can be selected.
  • one of the multiple boards provides service for new users, or updates the load of multiple boards and re-selects the board, or selects the lightest board in the first load level to provide new users. The service, or suspends the access request of the new user and waits for the board that meets the requirements of the new user to appear and then services the new user.
  • a load indicator required by the service when the new user requests the service for example, the file system input and output load of the board that requests the recording service requirement is large, and one or more boards in the first load level There may not be a board that can meet the load capacity of the demand, and in the one or more boards of the second load level, although the predetermined load indicator is higher than the first load level, the predetermined load indicator is If the load capacity of the new user is sufficient, the board may have a load capacity that meets the load requirements of the new user. Therefore, you can select a board as the new user in the second load level. Provide services.
  • the load level of the corresponding one of the plurality of boards may be updated in the case of at least one of the following: reaching a predetermined update period; the corresponding board starts to provide services for the new user; Among the users of the services provided by the board, the user is powered off; the corresponding board is powered off or an abnormality occurs.
  • the load of the corresponding board that determines the power-off or abnormality is the full load.
  • the present embodiment further provides a load balancing device, which is used to implement the foregoing load balancing method. The function implementation of the device has been described in the foregoing method embodiments, and details are not described herein.
  • 2 is a schematic structural diagram of a load balancing apparatus according to an embodiment of the present invention.
  • the apparatus includes a determining module 22 and a selecting module 24, wherein the determining module 22 is configured to allocate a plurality of boards according to a content distribution network.
  • the load is determined by determining a load level corresponding to each of the plurality of boards;
  • the selecting module 24 is coupled to the determining module 22, and is configured to select one of the one or more boards corresponding to the first load level in the load level.
  • the board provides services for new users, where the load of one or more boards corresponding to the first load level is lower than the load corresponding to other load levels in the load level.
  • the modules and units involved in the embodiments of the present invention may be implemented by software, or may be implemented by hardware.
  • the described modules and units in this embodiment may also be disposed in a processor.
  • a processor includes a determining module 22 and a selecting module 24. Where the names of these modules are in a certain In this case, the definition of the module itself is not limited.
  • the determining module 22 can also be described as: "Setting to determine the load corresponding to each of the plurality of boards according to the load of the plurality of boards of the content distribution network. Level module".
  • the apparatus further comprises: a second determining module coupled to the determining module 22, configured to determine the loads of the plurality of boards respectively according to the predetermined load indicator.
  • the predetermined load indicator includes, but is not limited to, at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board.
  • FIG. 3 is a schematic diagram of a preferred structure of a load balancing apparatus according to an embodiment of the present invention. As shown in FIG.
  • the selecting module 24 includes: a determining unit 242, configured to determine whether the load indicator of one or more boards meets The load indicator of the new user requirement, where the load indicator of the new user requirement includes at least one of the following: a file system input and output indicator of the board, a network port traffic indicator of the board, and a central processor load indicator of the board;
  • the board is coupled to the determining unit 242, and is configured to select a board in which one or more load indicators of the board meet the requirements of the new user to provide services for the new user.
  • the determining module 22 is further configured to map the plurality of boards into the plurality of queues according to the plurality of threshold ranges in which the loads of the plurality of boards respectively fall, wherein one of the plurality of queues respectively corresponds to A load level in the load level.
  • the determining unit 242 is further configured to sequentially determine whether the load indicator of one or more boards in the queue corresponding to the first load level meets the load indicator of the new user requirement, where the load indicator of the new user requirement includes at least the following The input and output indicators of the file system of the board, the network port traffic indicator of the board, and the CPU load indicator of the board; the selecting unit 244 is further configured to select a load indicator that satisfies the new user requirement in the queue corresponding to the first load level.
  • the first board provides services for new users.
  • FIG. 4 is a schematic diagram of a preferred structure of a load balancing apparatus according to an embodiment of the present invention. As shown in FIG.
  • the apparatus further includes: an updating module 42 configured to update multiple orders in the case of at least one of the following The load level of the corresponding board in the board: The predetermined update period is reached; the corresponding board starts to provide services for the new user; the user of the service provided by the corresponding board exits; the corresponding board is powered off or an abnormality occurs. Ensure that the load of the corresponding board that is powered off or abnormal is full.
  • the update module 42 can be located in the determination module 22, or coupled to the selection module 24, or coupled to the determination module 22 and the selection module 24, respectively. Description and description are made below in conjunction with the preferred embodiments.
  • the preferred embodiment provides a load balancing method based on hierarchical control in a CDN system to solve the problem of user load balancing in the CDN system and improve the stability of the system operation.
  • a mechanism for implementing load balancing by using a hierarchical control strategy according to a preferred embodiment of the present invention adopts the following solution: Step 1: Select an appropriate indicator to reflect the load of the board, and generally select an indicator that the board is in short supply, such as The file 10, the network port traffic, the CPU load, and the like are described in the preferred embodiment with the file 10 as an example.
  • the file 10 is calculated according to the service carried on the board, and the preemption mode is adopted.
  • Step 2 according to the real-time occupancy of the file 10, the board is divided into several levels (equivalent to the load level), such as 20% of the idle level is the lowest level, followed by idle 20% ⁇ 50%, 50% ⁇ 80%, 80% ⁇ 100% and so on.
  • Step 3 When the user (the signaling processing module or the load balancing module can be called a user) selects, the most idle level queue is selected first. If there are multiple boards in the most idle level queue, they are on the boards. Inter-round selection.
  • FIG. 5 is a schematic diagram of a functional module topology according to a preferred embodiment of the present invention.
  • the system for implementing the foregoing mechanism includes: a set top box, a signaling processing module, a load balancing module, and multiple streaming media.
  • the set-top box sends the signaling to the signaling processing module, and the signaling processing module sends the request to the load balancing module.
  • the load balancing module determines a board that provides the service, and finally forwards the signaling to the streaming media module on the board.
  • Conduct the service. 6 is a schematic flowchart of a hierarchical control process according to a preferred embodiment of the present invention. Referring to FIG.
  • Step 62 When the module is powered on, create multiple first-in-a-box for round selection.
  • the rank queues are assigned to the queues of the ranks according to the 10 occupations.
  • the boards with abnormal status or manually blocked are placed directly into the busy queue.
  • Step 64 When the user accesses, first selects in the most idle queue to determine whether the board meets the service conditions, such as the network port status, CPU, file 10, etc.
  • FIG. 7 is a schematic diagram of a rank queue according to a preferred embodiment of the present invention.
  • the MSTU GRAND X in the figure respectively represents four queues, and the number in each queue is the identification number of the corresponding board. It should be noted that the above preferred embodiment is described by taking only the file 10 as an example, and is not intended to limit the present invention.
  • the foregoing embodiment or the preferred embodiment may be implemented by using other load indicators or a combination of the load indicators as needed, and details are not described herein again. It can be seen from the above description that the present invention can combine the advantages of the round selection strategy and the preferred strategy through the above embodiments or the preferred embodiments, and on the one hand, the load balancing can be ensured as much as possible, and on the other hand, the maximum use of resources can be realized.
  • the advantages of the solution in the related art are as follows: 1.
  • the selected indicator can be the bottleneck of the board, or the combination of several indicators, so as to avoid a situation where a certain indicator is particularly high;
  • the load level corresponding to each of the plurality of boards is determined according to the load of the plurality of boards of the content distribution network; and the one corresponding to the first load level in the load level is selected or Multiple boards
  • a board provides services for new users, wherein the load of one or more boards corresponding to the first load level is lower than the load corresponding to other load levels in the load level, and the content distribution network in the related art is solved.
  • modules or steps of the present invention can be implemented by a general-purpose computing device, which can be concentrated on a single computing device or distributed over a network composed of multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device, such that they may be stored in the storage device by the computing device, or they may be separately fabricated into individual integrated circuit modules, or they may be Multiple modules or steps are made into a single integrated circuit module. Thus, the invention is not limited to any specific combination of hardware and software.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Information Transfer Between Computers (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明公开了一种负载均衡方法和装置,其中,该方法包括:根据内容分发网络的多个单板的负载,确定多个单板中每个单板对应的负载等级;选择负载等级中的第一负载等级对应的一个或多个单板中的一个单板为新用户提供服务,其中,第一负载等级对应的一个或多个单板的负载低于该负载等级中其他负载等级对应的负载。通过本发明,解决了相关技术中的内容分发网络的轮选策略或者优选策略导致的负载不均衡的问题,从而实现了内容分发网络的多个单板的负载均衡。

Description

负载均衡方法和装置 技术领域 本发明涉及通信领域, 具体而言, 涉及一种负载均衡方法和装置。 背景技术 在内容分发网络 (Content Delivery Network, 简称为 CDN) 系统中, 为了增加系 统总的服务能力, 往往需要多个单板为用户提供服务。 信令处理模块需要在用户接入 时, 选择为用户提供服务的单板, 出于流程的可控性, 这种选择往往是一次性的。 因 此, 如何选择为用户服务的单板, 并保证单板间负载均衡以减少单板负载过大的几率 就显得尤为重要。 相关技术中的一种 CDN系统的负载均衡方案采用的是轮选策略, 即在所有单板 中轮流选取服务的单板, 这种方法可以简单的实现用户数上的均衡, 且正常情况下较 为稳定, 不用提供外部数据 (例如单板的负载数据) 就能实现。 但是发明人发现在采 用轮选策略时, 以下几种情况下会出现问题:
1、 由于用户点播内容码率不一样, 虽然用户数均衡, 其他指标, 如文件输入输出 (10)、 网口流量等, 很容易出现不均衡;
2、系统中某些单板下电或者异常, 因为后续没有调整策略, 其他单板的压力会比 较大;
3、 频道录制等业务对单板的资源消耗较大, 如果和普通 VOD用户一样轮选, 尤 其单板复位导致频道集中, 这样频道多的单板压力会非常大; 4、 现网用户退出时机是随机的, 因此也会导致单板间的不均衡。 相关技术中的另一种 CDN系统的负载均衡方案采用的是针对单板负载指标的优 选策略, 即选择一种典型的指标, 如文件系统 10、 网口流量、 中央处理器 (Central Processing Unit, 简称为 CPU)等, 每次选择负载最轻的单板服务。 这种策略对系数的 选择要求特别高, 并且需要及时并且准确的数据支持。 并且, 发明人在研究过程中发 现, 由于该策略仅对于最优的一个单板进行选择, 在数据统计或者上报不够及时, 而 用户数快速上线或者迁移时, 容易导致用户集中在一个单板, 从而使得这块单板由于 负载突然增大导致该单板服务的所有用户出现卡顿。 针对相关技术中的内容分发网络的轮选策略或者优选策略导致的负载不均衡的问 题, 目前尚未提出有效的解决方案。 发明内容 本发明实施例提供了一种负载均衡方法和装置, 以至少解决相关技术中的内容分 发网络的轮选策略或者优选策略导致的负载不均衡的问题。 根据本发明实施例的一个方面, 提供了一种负载均衡方法, 包括: 根据内容分发 网络的多个单板的负载, 确定所述多个单板中每个单板对应的负载等级; 选择所述负 载等级中的第一负载等级对应的一个或多个单板中的一个单板为新用户提供服务, 其 中, 所述第一负载等级对应的所述一个或多个单板的所述负载低于所述负载等级中其 他负载等级对应的所述负载。 优选地,在确定所述多个单板中每个单板对应的负载等级之前,所述方法还包括: 根据预定负载指标, 分别确定所述多个单板的所述负载。 优选地, 所述预定负载指标包括以下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理器负载指标。 优选地, 选择所述第一负载等级对应的所述一个或多个单板中的所述一个单板包 括: 判断所述一个或多个单板的负载指标是否满足所述新用户需求的负载指标,其中, 所述新用户需求的负载指标包括以下至少之一: 单板的文件系统输入输出指标、 单板 的网口流量指标、 单板的中央处理器负载指标; 选择所述一个或多个单板中负载指标 满足所述新用户需求的负载指标的一个单板为所述新用户提供服务。 优选地, 确定所述多个单板中每个单板对应的所述负载等级包括: 根据所述多个 单板的所述负载分别落入的多个阈值范围, 将所述多个单板依次映射到多个队列中, 其中, 所述多个队列中的一个队列分别对应于所述负载等级中的一个负载等级。 优选地, 选择所述第一负载等级对应的所述一个或多个单板中的所述一个单板包 括: 依次判断所述第一负载等级对应的队列中的一个或多个单板的负载指标是否满足 所述新用户需求的负载指标, 其中, 所述新用户需求的负载指标包括以下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理器负载指标; 选择所述第一负载等级对应的队列中满足所述新用户需求的负载指标的第一个单板为 所述新用户提供服务。 优选地, 所述方法还包括: 在以下至少之一的情况下更新所述多个单板中的相应 单板的所述负载等级: 达到预定的更新周期;所述相应单板开始为新的用户提供服务; 所述相应单板所提供服务的用户中有用户退出; 所述相应单板下电或者出现异常, 其 中, 确定所述下电或者出现异常的所述相应单板的负载为满负载。 根据本发明实施例的另一个方面, 还提供了一种负载均衡装置, 包括: 确定模块, 设置为根据内容分发网络的多个单板的负载, 确定所述多个单板中每个单板对应的负 载等级; 选择模块, 设置为选择所述负载等级中的第一负载等级对应的一个或多个单 板中的一个单板为新用户提供服务, 其中, 所述第一负载等级对应的所述一个或多个 单板的所述负载低于所述负载等级中其他负载等级对应的所述负载。 优选地, 所述选择模块包括: 判断单元, 设置为判断所述一个或多个单板的负载 指标是否满足所述新用户需求的负载指标, 其中, 所述新用户需求的负载指标包括以 下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理 器负载指标; 选择单元, 设置为选择所述一个或多个单板中负载指标满足所述新用户 需求的负载指标的一个单板为所述新用户提供服务。 优选地, 所述装置还包括: 更新模块, 设置为在以下至少之一的情况下更新所述 多个单板中的相应单板的所述负载等级: 达到预定的更新周期; 所述相应单板开始为 新的用户提供服务; 所述相应单板所提供服务的用户中有用户退出; 所述相应单板下 电或者出现异常,其中,确定所述下电或者出现异常的所述相应单板的负载为满负载。 通过本发明实施例, 采用根据内容分发网络的多个单板的负载, 确定多个单板中 每个单板对应的负载等级; 选择负载等级中的第一负载等级对应的一个或多个单板中 的一个单板为新用户提供服务, 其中, 第一负载等级对应的一个或多个单板的负载低 于该负载等级中其他负载等级对应的负载的方式, 解决了相关技术中的内容分发网络 的轮选策略或者优选策略导致的负载不均衡的问题, 从而实现了内容分发网络的多个 单板的负载均衡。 附图说明 此处所说明的附图用来提供对本发明的进一步理解, 构成本申请的一部分, 本发 明的示意性实施例及其说明用于解释本发明, 并不构成对本发明的不当限定。 在附图 中: 图 1是根据本发明实施例的负载均衡方法的流程示意图; 图 2是根据本发明实施例的负载均衡装置的结构示意图; 图 3是根据本发明实施例的负载均衡装置的优选结构示意图一; 图 4是根据本发明实施例的负载均衡装置的优选结构示意图二; 图 5是根据本发明优选实施例的功能模块拓扑结构示意图; 图 6是根据本发明优选实施例的分级控制过程的流程示意图; 图 7是根据本发明优选实施例的一个等级队列的示意图。 具体实施方式 需要说明的是, 在不冲突的情况下, 本申请中的实施例及实施例中的特征可以相 互组合。 下面将参考附图并结合实施例来详细说明本发明。 在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执 行, 并且, 虽然在流程图中示出了逻辑顺序, 但是在某些情况下, 可以以不同于此处 的顺序执行所示出或描述的步骤。 本实施例提供了一种负载均衡方法, 图 1是根据本发明实施例的负载均衡方法的 流程示意图, 如图 1所示, 该方法包括如下流程: 步骤 S102, 根据内容分发网络的多个单板的负载, 确定多个单板中每个单板对应 的负载等级; 步骤 S104,选择负载等级中的第一负载等级对应的一个或多个单板中的一个单板 为新用户提供服务, 其中, 第一负载等级对应的一个或多个单板的负载低于负载等级 中其他负载等级对应的负载。 通过上述步骤,采用根据多个单板的负载分别确定其中每个单板对应的负载等级, 并选择负载等级表示为负载低的第一负载等级中的一个或多个单板中的一个单板为新 用户提供服务。 上述步骤所示的方案是基于单板负载的均衡策略, 因此相对于相关技 术中的轮选策略, 该方案能够实现负载的均衡; 由于上述方案是基于同一个负载等级 中的一个或多个单板的均衡策略, 因此相对于相关技术中的优选策略, 该方案对系数 的要求和数据的及时、 准确性相对要低, 而且不容易出现之前负载最轻的单个单板的 负载突然增大的问题。 由此可见, 通过本实施例的上述步骤, 解决了相关技术中的内 容分发网络的轮选策略或者优选策略导致的负载不均衡的问题, 从而实现了内容分发 网络的多个单板的负载均衡。 优选地, 多个单板中每个单板的负载是根据预定负载指标确定的, 例如, 预定负 载指标可以是以下至少之一的指标: 单板的文件系统输入输出指标、 单板的网口流量 指标、 单板的中央处理器负载指标。 当然, 预定负载指标不限于上述的指标, 用户可 以根据需要或者系统的要求设置其他的负载指标或者对这些负载指标进行单独选择或 组合应用。 在用户选取负载指标时, 优选地选择对于这类单板而言容易造成整体负载瓶颈的 负载指标, 例如: 对已一类单板而言, 其中央处理器处理能力相对较强不容易出现指 标过高的情况, 而造成其负载过大的主要因素往往是网口流量指标过高, 此时可以优 选包含网口流量指标作为预定负载指标。 优选地, 可以对多个单板的多种负载指标进行归一化处理, 并在同一个内容分发 网络的负载均衡处理中对不同单板或者不同类型的单板分别采用归一化处理后的不同 预定负载指标来确定各自单板的负载。 优选地, 不同单板或者不同类型的单板分别选 择容易造成其整体负载瓶颈的负载指标作为该预定负载指标。 优选地,在步骤 S104中,为了使得待加入的新用户在选择一个单板为其提供服务 后, 被选择的单板不会出现负载过大的问题, 可以在第一负载等级对应的一个或多个 单板中选择单板时先判断这些单板的负载指标是否满足新用户需求的负载指标,例如, 剩余的负载能力是否满足新用户请求的服务所需要的负载能力, 从而选择这一个或多 个单板中负载指标满足新用户需求的负载指标的一个单板为新用户提供服务。 优选地, 新用户需求的负载指标包括但不限于以下至少之一: 单板的文件系统输 入输出指标、 单板的网口流量指标、 单板的中央处理器负载指标。 在步骤 S102中,可以根据多个单板的负载分别落入的多个阈值范围,将多个单板 依次映射到多个队列中, 其中, 多个队列中的一个队列分别对应于负载等级中的一个 负载等级。 其中, 多个阈值范围中的一个阈值范围对应一个负载等级, 该阈值范围或 负载等级的个数可以由用户根据需求的负载均衡的粒度和精度进行选择, 阈值范围或 负载等级的个数越多, 则该负载均衡方法的负载均衡能力越强,但需求的数据实时性、 准确性需求将越大。 优选地,在使用上述的队列策略的情况下,在步骤 S104中可以依次判断第一负载 等级对应的队列中的一个或多个单板的负载指标是否满足新用户需求的负载指标, 并 选择第一负载等级对应的队列中满足新用户需求的负载指标的第一个单板为新用户提 供服务。 其中, 新用户需求的负载指标包括以下至少之一: 单板的文件系统输入输出 指标、 单板的网口流量指标、 单板的中央处理器负载指标; 优选地, 在判断到第一负载等级的一个或多个单板中没有能够满足新用户需求的 负载指标的第一个单板可以为新用户提供服务的情况下, 可以选择负载比第一负载等 级较高的第二负载等级的一个或多个单板中的一个单板为新用户提供服务, 或者更新 多个单板的负载情况并重新进行单板的选择, 或者选择第一负载等级中负载最轻的单 板为新用户提供服务, 或者挂起新用户的接入请求并等待符合该新用户需求的单板出 现后再为该新用户提供服务。 其中, 由于新用户请求服务时, 该服务所需求的某一个负载指标: 例如请求录制 业务需求的单板的文件系统输入输出负载较大, 而在第一负载等级中的一个或多个单 板中可能并不存在能够满足其需求的负载能力的单板, 而在第二负载等级中的一个或 多个单板中, 虽然其预定负载指标较第一负载等级较高, 但是在预定负载指标能够满 足新用户需求的负载能力的情况下, 可能也会存在单板的文件系统输入输出负载能力 满足新用户需求的负载能力的单板, 因此可以在第二负载等级中选择单板为新用户提 供服务。 优选地, 在本实施例中, 可以在以下至少之一的情况下更新多个单板中的相应单 板的负载等级: 达到预定的更新周期; 相应单板开始为新的用户提供服务; 相应单板 所提供服务的用户中有用户退出; 相应单板下电或者出现异常, 其中, 确定下电或者 出现异常的相应单板的负载为满负载。 本实施例还提供了一种负载均衡装置, 该装置用于实现上述负载均衡方法, 该装 置的功能实现已经在上述方法实施例中进行了说明, 在此不再赘述。 图 2是根据本发明实施例的负载均衡装置的结构示意图, 如图 2所示, 该装置包 括确定模块 22和选择模块 24, 其中, 确定模块 22, 设置为根据内容分发网络的多个 单板的负载,确定多个单板中每个单板对应的负载等级;选择模块 24耦合至确定模块 22, 设置为选择负载等级中的第一负载等级对应的一个或多个单板中的一个单板为新 用户提供服务, 其中, 第一负载等级对应的一个或多个单板的负载低于负载等级中其 他负载等级对应的负载。 本发明的实施例中所涉及到的模块、 单元可以通过软件的方式实现, 也可以通过 硬件的方式来实现。本实施例中的所描述的模块、单元也可以设置在处理器中, 例如, 可以描述为: 一种处理器包括确定模块 22、 选择模块 24。 其中, 这些模块的名称在某 种情况下并不构成对该模块本身的限定, 例如, 确定模块 22还可以被描述为 "设置为 根据内容分发网络的多个单板的负载确定多个单板中每个单板对应的负载等级的模 块"。 优选地, 该装置还包括: 第二确定模块耦合至确定模块 22, 设置为根据预定负载 指标, 分别确定多个单板的负载。 优选地, 该预定负载指标包括但不限于以下至少之一: 单板的文件系统输入输出 指标、 单板的网口流量指标、 单板的中央处理器负载指标。 图 3是根据本发明实施例的负载均衡装置的优选结构示意图一, 如图 3所示, 优 选地, 选择模块 24包括: 判断单元 242, 设置为判断一个或多个单板的负载指标是否 满足新用户需求的负载指标, 其中, 新用户需求的负载指标包括以下至少之一: 单板 的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理器负载指标; 选择 单元 244耦合至判断单元 242, 设置为选择一个或多个单板中负载指标满足新用户需 求的负载指标的一个单板为新用户提供服务。 优选地,确定模块 22还设置为根据多个单板的负载分别落入的多个阈值范围,将 多个单板依次映射到多个队列中, 其中, 多个队列中的一个队列分别对应于负载等级 中的一个负载等级。 优选地, 判断单元 242还设置为依次判断第一负载等级对应的队列中的一个或多 个单板的负载指标是否满足新用户需求的负载指标, 其中, 新用户需求的负载指标包 括以下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央 处理器负载指标; 选择单元 244还设置为选择第一负载等级对应的队列中满足新用户 需求的负载指标的第一个单板为新用户提供服务。 图 4是根据本发明实施例的负载均衡装置的优选结构示意图二, 如图 4所示, 优 选地, 该装置还包括: 更新模块 42, 设置为在以下至少之一的情况下更新多个单板中 的相应单板的负载等级: 达到预定的更新周期; 相应单板开始为新的用户提供服务; 相应单板所提供服务的用户中有用户退出; 相应单板下电或者出现异常, 其中, 确定 下电或者出现异常的相应单板的负载为满负载。其中, 更新模块 42可以位于确定模块 22中, 也可以与选择模块 24耦合, 或者与确定模块 22和选择模块 24分别耦合。 下面结合优选实施例进行描述和说明。 本优选实施例提供了一种 CDN 系统中基于分级控制的负载均衡方法, 以解决 CDN系统中用户负载均衡问题, 提高系统运行的稳定性。 本发明优选实施例设计的一种利用分级控制策略实现负载均衡的机制采用如下的 方案: 步骤 1, 选择一个合适的指标用来反应单板的负载情况, 一般是选择单板紧缺的 指标, 如文件 10、 网口流量、 CPU负载等, 在本优选实施例中以文件 10为例进行说 明。 文件 10是根据单板上承载的业务计算而来, 采用预占的方式, 如 VOD用户占用 1倍码率的 10, 频道录制占用 5倍码率的 10。 对于 10, 可以根据需求和性能测试, 提供一个单板能够承受的最大值。 步骤 2, 根据文件 10实时占用情况将单板分成几个等级 (相当于负载等级), 如 空闲 20%以下属于最低的等级, 依次是空闲 20%~50%, 50%~80%, 80%~100%等。 步骤 3, 用户 (信令处理模块或负载均衡模块等都可以称之为用户)选择的时候, 首先选择最空闲的等级队列, 如果最空闲等级队列存在多个单板时, 就在这些单板间 轮选。 下面结合附图对本优选实施例进行说明。 图 5是根据本发明优选实施例的功能模块拓扑结构示意图, 如图 5所示, 用于实 现上述机制的系统包括: 机顶盒、 信令处理模块、 负载均衡模块、 多个流媒体。 其中, 机顶盒将信令发给信令处理模块, 信令处理模块发送请求到负载均衡模块, 由负载均 衡模块确定一个提供服务的单板, 最后将信令转发给这个单板上的流媒体模块进行服 务。 图 6是根据本发明优选实施例的分级控制过程的流程示意图, 参考图 6所示, 该 过程的实现包括的步骤如下: 步骤 62, 模块上电时, 创建多个用于轮选的先入先出的等级队列, 根据 10占用 情况, 将所有单板分配到各个等级队列中, 其中, 状态不正常或者人工闭塞的单板直 接放入最忙队列中; 步骤 64, 用户接入时, 先在最空闲队列中轮选, 判断单板是否符合服合条件, 如 网口状态, CPU, 文件 10等, 不符合条件则选择队列中下一个单元(即队列中的下一 个单板); 步骤 66, 选中某单板后, 对 10占用情况进行更新, 并重新评估单板应属的队列, 将其放入该队列的末尾; 步骤 68, 用户退出时, 重新评估是否需要更新等级队列。 图 7是根据本发明优选实施例的一个等级队列的示意图,图中的 MSTU GRAND X 分别表示 4个队列, 每个队列中的数字为相应单板的标识号。 需要说明的是,上述优选实施例仅以文件 10为例进行说明,并不用于对本发明的 任何限制。 根据需要可以选择其他负载指标或者这些负载指标的组合来实现上述实施 例或优选实施例, 在此不再赘述。 由上述描述可知, 本发明通过上述实施例或优选实施例, 可以很好的结合了轮选 策略和优选策略的优点, 一方面尽量保证负载均衡, 另一方面也可以实现资源的最大 使用, 相对于相关技术中的方案, 优点详述如下: 1、选择的指标可以是单板的瓶颈所在, 也可以上几种指标的综合, 这样尽量避免 出现某一种指标特别高的情况;
2、如果某块单板异常复位之后, 可以确定其处于最空闲队列, 这样开始阶段就会 尽量选在该块单板, 直到等级队列发生变化;
3、如果某块单板承载的频道很多, 也会在选择的指标中有所体现, 这样会将其放 入较忙的队列中, 在有选择的情况下, 尽量不会再加用户上去;
4、虽然用户退出时机是随机的, 但是只要等级队列发生变化, 就会对后续用户分 布进行调整, 保证不会出现大的不均衡;
5、 由于在等级队列内仍然采用轮选, 因此对用户上线速度并不是很敏感, 也即即 使用户上线速度比较集中, 也不容易出现同一单板的负载突然增大的问题。 工业实用性 通过本发明实施例, 采用根据内容分发网络的多个单板的负载, 确定多个单板中 每个单板对应的负载等级; 选择负载等级中的第一负载等级对应的一个或多个单板中 的一个单板为新用户提供服务, 其中, 第一负载等级对应的一个或多个单板的负载低 于该负载等级中其他负载等级对应的负载的方式, 解决了相关技术中的内容分发网络 的轮选策略或者优选策略导致的负载不均衡的问题, 从而实现了内容分发网络的多个 单板的负载均衡。 显然, 本领域的技术人员应该明白, 上述的本发明的各模块或各步骤可以用通用 的计算装置来实现, 它们可以集中在单个的计算装置上, 或者分布在多个计算装置所 组成的网络上, 可选地, 它们可以用计算装置可执行的程序代码来实现, 从而, 可以 将它们存储在存储装置中由计算装置来执行, 或者将它们分别制作成各个集成电路模 块, 或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。 这样, 本发明 不限制于任何特定的硬件和软件结合。 以上上述仅为本发明的优选实施例而已, 并不用于限制本发明, 对于本领域的技 术人员来说, 本发明可以有各种更改和变化。 凡在本发明的精神和原则之内, 所作的 任何修改、 等同替换、 改进等, 均应包含在本发明的保护范围之内。

Claims

权 利 要 求 书
1. 一种负载均衡方法, 包括:
根据内容分发网络的多个单板的负载, 确定所述多个单板中每个单板对应 的负载等级;
选择所述负载等级中的第一负载等级对应的一个或多个单板中的一个单板 为新用户提供服务, 其中, 所述第一负载等级对应的所述一个或多个单板的所 述负载低于所述负载等级中其他负载等级对应的所述负载。
2. 根据权利要求 1所述的方法, 其中, 在确定所述多个单板中每个单板对应的负 载等级之前, 所述方法还包括:
根据预定负载指标, 分别确定所述多个单板的所述负载。
3. 根据权利要求 2所述的方法, 其中, 所述预定负载指标包括以下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理器 负载指标。
4. 根据权利要求 1所述的方法, 其中, 选择所述第一负载等级对应的所述一个或 多个单板中的所述一个单板包括: 判断所述一个或多个单板的负载指标是否满足所述新用户需求的负载指 标, 其中, 所述新用户需求的负载指标包括以下至少之一: 单板的文件系统输 入输出指标、 单板的网口流量指标、 单板的中央处理器负载指标;
选择所述一个或多个单板中负载指标满足所述新用户需求的负载指标的一 个单板为所述新用户提供服务。
5. 根据权利要求 1所述的方法, 其中, 确定所述多个单板中每个单板对应的所述 负载等级包括:
根据所述多个单板的所述负载分别落入的多个阈值范围, 将所述多个单板 依次映射到多个队列中, 其中, 所述多个队列中的一个队列分别对应于所述负 载等级中的一个负载等级。
6. 根据权利要求 5所述的方法, 其中, 选择所述第一负载等级对应的所述一个或 多个单板中的所述一个单板包括: 依次判断所述第一负载等级对应的队列中的一个或多个单板的负载指标是 否满足所述新用户需求的负载指标, 其中, 所述新用户需求的负载指标包括以 下至少之一: 单板的文件系统输入输出指标、 单板的网口流量指标、 单板的中 央处理器负载指标;
选择所述第一负载等级对应的队列中满足所述新用户需求的负载指标的第 一个单板为所述新用户提供服务。 根据权利要求 1至 6中任一项所述的方法, 其中, 所述方法还包括: 在以下至 少之一的情况下更新所述多个单板中的相应单板的所述负载等级: 达到预定的 更新周期; 所述相应单板开始为新的用户提供服务; 所述相应单板所提供服务 的用户中有用户退出; 所述相应单板下电或者出现异常, 其中, 确定所述下电 或者出现异常的所述相应单板的负载为满负载。 一种负载均衡装置, 包括:
确定模块, 设置为根据内容分发网络的多个单板的负载, 确定所述多个单 板中每个单板对应的负载等级; 选择模块, 设置为选择所述负载等级中的第一负载等级对应的一个或多个 单板中的一个单板为新用户提供服务, 其中, 所述第一负载等级对应的所述一 个或多个单板的所述负载低于所述负载等级中其他负载等级对应的所述负载。 根据权利要求 8所述的装置, 其中, 所述选择模块包括:
判断单元, 设置为判断所述一个或多个单板的负载指标是否满足所述新用 户需求的负载指标, 其中, 所述新用户需求的负载指标包括以下至少之一: 单 板的文件系统输入输出指标、 单板的网口流量指标、 单板的中央处理器负载指 标;
选择单元, 设置为选择所述一个或多个单板中负载指标满足所述新用户需 求的负载指标的一个单板为所述新用户提供服务。 根据权利要求 8或 9所述的装置, 其中, 所述装置还包括:
更新模块, 设置为在以下至少之一的情况下更新所述多个单板中的相应单 板的所述负载等级: 达到预定的更新周期; 所述相应单板开始为新的用户提供 服务; 所述相应单板所提供服务的用户中有用户退出; 所述相应单板下电或者 出现异常,其中,确定所述下电或者出现异常的所述相应单板的负载为满负载。
PCT/CN2014/079002 2013-12-06 2014-05-30 负载均衡方法和装置 WO2014180418A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310658696.2 2013-12-06
CN201310658696.2A CN104702521A (zh) 2013-12-06 2013-12-06 负载均衡方法和装置

Publications (1)

Publication Number Publication Date
WO2014180418A1 true WO2014180418A1 (zh) 2014-11-13

Family

ID=51866798

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/079002 WO2014180418A1 (zh) 2013-12-06 2014-05-30 负载均衡方法和装置

Country Status (2)

Country Link
CN (1) CN104702521A (zh)
WO (1) WO2014180418A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113767597A (zh) * 2020-04-03 2021-12-07 华为技术有限公司 用于基于周期的负载均衡的网络设备、系统和方法
CN113923166A (zh) * 2021-10-22 2022-01-11 北京字跳网络技术有限公司 负载均衡方法和分布式集群系统

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106998340B (zh) * 2016-01-22 2020-10-30 大唐移动通信设备有限公司 一种板卡资源的负载均衡方法及装置
CN108289113A (zh) * 2017-01-10 2018-07-17 中兴通讯股份有限公司 vBNG系统多虚拟机负荷分担方法及vBNG系统设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1509009A (zh) * 2002-12-19 2004-06-30 华为技术有限公司 一种业务资源的分配方法
CN1592166A (zh) * 2003-09-05 2005-03-09 深圳市中兴通讯股份有限公司南京分公司 基于wcdma系统的多用户类型分布式环境下的接纳控制方法
CN102076020A (zh) * 2011-01-25 2011-05-25 中兴通讯股份有限公司 一种se资源的分配方法及装置
CN102870374A (zh) * 2012-06-21 2013-01-09 华为技术有限公司 负荷分担方法及装置、单板

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1509009A (zh) * 2002-12-19 2004-06-30 华为技术有限公司 一种业务资源的分配方法
CN1592166A (zh) * 2003-09-05 2005-03-09 深圳市中兴通讯股份有限公司南京分公司 基于wcdma系统的多用户类型分布式环境下的接纳控制方法
CN102076020A (zh) * 2011-01-25 2011-05-25 中兴通讯股份有限公司 一种se资源的分配方法及装置
CN102870374A (zh) * 2012-06-21 2013-01-09 华为技术有限公司 负荷分担方法及装置、单板

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113767597A (zh) * 2020-04-03 2021-12-07 华为技术有限公司 用于基于周期的负载均衡的网络设备、系统和方法
CN113923166A (zh) * 2021-10-22 2022-01-11 北京字跳网络技术有限公司 负载均衡方法和分布式集群系统
CN113923166B (zh) * 2021-10-22 2024-02-23 北京字跳网络技术有限公司 负载均衡方法和分布式集群系统

Also Published As

Publication number Publication date
CN104702521A (zh) 2015-06-10

Similar Documents

Publication Publication Date Title
US11888756B2 (en) Software load balancer to maximize utilization
US7712103B2 (en) Method and system for managing load balancing in data processing system
US10523732B2 (en) Multi-CDN digital content streaming
Hong et al. Achieving high utilization with software-driven WAN
US9389920B2 (en) Intelligent data center cluster selection
WO2021126295A1 (en) Request throttling in distributed storage systems
US9148495B2 (en) Dynamic runtime choosing of processing communication methods
US20160378570A1 (en) Techniques for Offloading Computational Tasks between Nodes
CN109120715A (zh) 一种云环境下动态负载均衡方法
WO2010100859A1 (ja) 分散システム
US20140143431A1 (en) Multi-cdn digital content streaming
EP2725862A1 (en) Resource allocation method and resource management platform
CN103986766B (zh) 自适应负载均衡作业任务调度方法及装置
US10419370B2 (en) Hierarchical packet buffer system
WO2014180418A1 (zh) 负载均衡方法和装置
WO2011088767A1 (zh) 内容分发的方法、系统及调度服务器
US9712580B2 (en) Pipelining for parallel network connections to transmit a digital content stream
WO2021164164A1 (zh) 一种存储服务质量控制方法、装置、设备及存储介质
WO2023103342A1 (zh) 集群资源配额分配方法、装置及电子设备
US20150253837A1 (en) Software Enabled Network Storage Accelerator (SENSA) - Power Savings in Arrays of Multiple RISC Cores
JP2010108300A (ja) 情報処理システム、及び情報処理システムにおけるi/oのパスへの割り当て方法
JP2001202318A (ja) データ配信システム
US8843671B1 (en) Dynamic resource allocation for serial attached SCSI devices
US10963305B2 (en) Low latency distributed counters for quotas
CN116909953A (zh) 一种多路径规划方法、装置及介质

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: 14794886

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14794886

Country of ref document: EP

Kind code of ref document: A1