WO2021248917A1 - 一种基于混合链的数据中心网络故障诊断和自动配置方法 - Google Patents

一种基于混合链的数据中心网络故障诊断和自动配置方法 Download PDF

Info

Publication number
WO2021248917A1
WO2021248917A1 PCT/CN2021/073501 CN2021073501W WO2021248917A1 WO 2021248917 A1 WO2021248917 A1 WO 2021248917A1 CN 2021073501 W CN2021073501 W CN 2021073501W WO 2021248917 A1 WO2021248917 A1 WO 2021248917A1
Authority
WO
WIPO (PCT)
Prior art keywords
data center
chain
center network
transaction
fault
Prior art date
Application number
PCT/CN2021/073501
Other languages
English (en)
French (fr)
Inventor
黄海平
陈雨昊
陈龙
肖甫
汪文明
朱洁
马子洋
李琦
Original Assignee
南京邮电大学
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 南京邮电大学 filed Critical 南京邮电大学
Publication of WO2021248917A1 publication Critical patent/WO2021248917A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • 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
    • H04L41/04—Network management architectures or arrangements
    • 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
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/04—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
    • H04L63/0407—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the identity of one or more communicating identities is hidden
    • H04L63/0421—Anonymous communication, i.e. the party's identifiers are hidden from the other party or parties, e.g. using an anonymizer
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/12—Applying verification of the received information
    • H04L63/123—Applying verification of the received information received data contents, e.g. message integrity

Definitions

  • the invention relates to the cross-technology field of host security and blockchain, in particular to a data center network fault diagnosis and automatic configuration method based on a hybrid chain.
  • each data center network with interdependent data is no longer separate from each other, but forms an alliance group.
  • the data of a mass commodity may be migrated from the data center network A in the production link to the data center network B in the sales link.
  • This type of data center network group poses a huge challenge to the centralized management of the data flow of each data center network.
  • the new server-centric data center network (group) structure such as BCube, DCell, BCCC, RCube, and GBC3, should have two major functions such as accurate fault identification and automatic system reconfiguration: the role of the former is Find and accurately locate faulty nodes in the data center network in real time as much as possible. The role of the latter is to notify the system administrator of the data center network to replace the faulty server with the backup server or to assign the data task of the faulty server to other servers, and Inform other neighboring data center networks that have data dependencies.
  • the former is usually implemented by using a system-level diagnosis model (such as a PMC model) combined with t diagnosis or t/t diagnosis technology.
  • a system-level diagnosis model such as a PMC model
  • t diagnosis or t/t diagnosis technology can be achieved by combining the fault log with the predetermined fault handling method.
  • these traditional processing technologies require a centralized processing mechanism, which is not perfect in terms of processing effects such as fault records, reliability, and evidence traceability, especially for ultra-large-scale data center network groups.
  • the purpose of the present invention is to provide a data center network fault diagnosis and automatic configuration method based on a hybrid chain.
  • the super-large-scale data center network group is regarded as a data-dependent alliance chain, and each independent data center network It is a member node in the consortium chain, and at the same time it acts as an independent private chain, in which servers or ordinary switching devices are used as member nodes in the private chain, thus forming a chain-in-chain of the data center network group.
  • the present invention provides a data center network fault diagnosis and automatic configuration method based on a hybrid chain.
  • the method includes the following steps:
  • Step 1 Establish a data center network architecture based on a hybrid chain: design an independent data center network architecture into a private chain, and the servers in the data center network are members of the private chain, and each member initializes its neighbor list; each has a data dependency relationship The adjacent data center network as members form a consortium chain, and at the same time form a data center network group, each member initializes its neighbor list; several private chains and consortium chains together form a hybrid blockchain;
  • Step 2 Initialization of the security parameters of the blockchain: an independent data center network is used as a private chain. Based on a certain public key cryptosystem, each member obtains its public-private key pair; the private chain uses a Merkle tree to store the transaction information of each member. Hope-valued path; after signing the root of the Merkle tree, it is stored in the private chain block and the integrity of the transaction information can be verified; each member conducts transactions through pseudonymized addresses; each adjacent data center network with data dependencies A consortium chain is formed.
  • each member obtains its public-private key pair in the consortium chain; the consortium chain uses a Merkle tree to store the transaction information hash value path of each member; sign the root of the Merkle tree It can then be stored in the alliance chain block and the integrity of the transaction information can be verified; each member also conducts transactions through pseudonymous addresses;
  • Step 3 Independent data center network completes fault node diagnosis and system automatic configuration: first determine the search strategy, traverse the entire data center network members; second, complete the fault mutual diagnosis test with neighbor members, and publish the fault mutual diagnosis information as a transaction on the private In the chain; then, according to the transaction information in the private chain, based on a certain diagnostic system model and diagnostic technology, the faulty members are put into the fault set, and the normal members are put into the normal set; finally, the members of the fault set are completed through the smart contract
  • the server processes and publishes processing transaction announcements;
  • Step 4 There may be data transmission dependence between the independent data center networks in the alliance chain. Two adjacent independent data center networks determine whether the currently transmitted data has a data dependence relationship through their respective data attribute sets. If there is data dependence, Relationship, record the transaction and complete the data transmission;
  • Step 5 Realize verification and tracking of all transaction records that have been published by the data center network group.
  • a further improvement lies in the establishment of the structure of the private chain and consortium chain in the hybrid chain in step 1:
  • the transaction information between members is mainly fault mutual diagnosis information, server replacement information or task adjustment information, mainly with the help of
  • the smart contract deployed in the private chain completes decentralized transaction information release and automatic configuration;
  • the transaction information between members is mainly connectivity test information or data dependent transmission information, mainly by means of deployment in the alliance chain
  • the smart contract completes the decentralized transaction information release and automatic configuration.
  • each independent data center network selects a search strategy according to actual needs: depth-first traversal strategy, breadth-first traversal first-in-first-out strategy, breadth-first traversal first-in-last-out strategy, and breadth-first traversal Priority queue strategy.
  • a further improvement is that: in the step 3, when completing the mutual fault diagnosis test with neighbor members, the mutual fault diagnosis information is formed into a transaction and released in the private chain, including the following steps:
  • Step 3.1 Each independent data center network randomly assigns a certain original node x, constructs its neighbor list N(x) according to the determined search strategy, and conducts fault mutual diagnosis test with neighbor node y in the neighbor list in turn.
  • the results of the diagnostic test are classified as and Four situations, where T means normal, F means fault, and T/F means unable to determine;
  • Step 3.2 The fault diagnosis result is used as the content of the transaction.
  • X and y are the initiators of the transaction at the same time. They respectively use the pseudonyms related to their public key information, and use their private keys to complete the digital signature of the transaction, and then publish the transaction on In the private chain, all members will receive the transaction information;
  • Step 3.3 The consensus node in the private chain finally records the transaction of the fault diagnosis result in the generated block through a certain consensus algorithm
  • Step 3.4 The member server that has completed the fault diagnosis will be marked. According to the search strategy, select another member server that has not been diagnosed or has not yet been diagnosed. Repeat steps 3.1-3.3 until all member servers in the independent data center network have been completed The fault is diagnosed or marked.
  • a further improvement is that: in the step 3, according to the transaction information recorded in the private chain, the smart contract is based on a certain diagnostic system model and diagnostic technology, and the faulty server is placed in the fault collection, and the normal server is placed in the normal collection.
  • smart contracts are divided into the following two situations: Optimistic diagnosis: Combining the diagnosis results of all neighbor servers in the data center network to determine and investigate one by one, and each server is classified into the fault set and the normal set. It is accurate; pessimistic diagnosis: the number of times the diagnosis result T of each server is more than the number of F is regarded as a normal server, and the number of times the diagnosis result F of each server is more than the number of T is regarded as a faulty server. If the diagnosis result of the server is only T/F, it is considered as a faulty server.
  • step 3 the processing of the member servers in the faulty set is completed through the smart contract, the backup server is used to replace the original faulty server or the tasks of the original faulty server are migrated to other normal servers, and the processing announcement is issued to the data Central network.
  • recording transactions between independent data center networks that may be dependent on data transmission in the alliance chain and completing data transmission includes the following steps:
  • Step 4.2 Test the connectivity between data center network A and data center network B; if connected, data center network A initiates a transaction, uses the transmitted data feature as the transaction content, uses the pseudonym related to its public key information, and uses its private The key completes the digital signature of the transaction and sends it to data center network B; if it is not connected, submit a fault report to the smart contract;
  • Step 4.3 The consensus node in the alliance chain finally records the transaction in the generated block through a certain consensus algorithm
  • Step 4.4 Data center network A and B complete data transmission.
  • a further improvement is that in the fifth step, for the alliance chain, the published transaction content connectivity test information or data dependent transmission information can be verified and tracked through the Merkle tree and the block chain respectively; for the private chain, the published transaction The content fault mutual diagnosis information, server replacement information or task adjustment information can be verified and tracked through the Merkle tree and the blocks of the blockchain respectively.
  • the beneficial effects of the present invention apply blockchain technology to a server-centric new data center network structure, and use smart contracts to realize decentralized management and automatic system configuration; based on a hybrid blockchain structure, independent
  • the data center network adopts a private chain structure. Unauthorized internal nodes cannot join the private chain.
  • the fault diagnosis results are recorded in the private chain in the form of transactions.
  • Each independent central network that may have data dependence forms a consortium chain.
  • Fig. 1 is a diagram of the data center network group structure of the present invention.
  • Figure 2 is a block structure diagram of the blockchain of the present invention.
  • Fig. 3 is a flowchart of the execution of the steps of the present invention.
  • this embodiment provides a data center network fault diagnosis and automatic configuration method based on a hybrid chain.
  • the method includes the following steps:
  • Step 1 Establish a data center network architecture based on a hybrid chain: design an independent data center network architecture into a private chain, and the servers in the data center network are members of the private chain, and each member initializes its neighbor list; each has a data dependency relationship The adjacent data center network as members form a consortium chain, and at the same time form a data center network group, each member initializes its neighbor list; several private chains and consortium chains together form a hybrid blockchain;
  • Step 2 Initialization of the security parameters of the blockchain: an independent data center network is used as a private chain. Based on a certain public key cryptosystem, each member obtains its public-private key pair; the private chain uses a Merkle tree to store the transaction information of each member. Hope-valued path; after signing the root of the Merkle tree, it is stored in the private chain block and the integrity of the transaction information can be verified; each member conducts transactions through pseudonymized addresses; each adjacent data center network with data dependencies A consortium chain is formed.
  • each member obtains its public-private key pair in the consortium chain; the consortium chain uses a Merkle tree to store the transaction information hash value path of each member; sign the root of the Merkle tree It can then be stored in the alliance chain block and the integrity of the transaction information can be verified; each member also conducts transactions through pseudonymous addresses;
  • Step 3 Independent data center network completes fault node diagnosis and system automatic configuration: first determine the search strategy, traverse the entire data center network members; second, complete the fault mutual diagnosis test with neighbor members, and publish the fault mutual diagnosis information as a transaction on the private In the chain; then, according to the transaction information in the private chain, based on a certain diagnostic system model and diagnostic technology, the faulty members are put into the fault set, and the normal members are put into the normal set; finally, the members of the fault set are completed through the smart contract
  • the server processes and publishes processing transaction announcements;
  • Step 4 There may be data transmission dependence between the independent data center networks in the alliance chain. Two adjacent independent data center networks determine whether the currently transmitted data has a data dependence relationship through their respective data attribute sets. If there is data dependence, Relationship, record the transaction and complete the data transmission;
  • Step 5 Realize verification and tracking of all transaction records that have been published by the data center network group.
  • the architecture of the private chain and consortium chain in the hybrid chain in the step 1 is established: in the private chain, the transaction information between members is mainly fault mutual diagnosis information, server replacement information or task adjustment information, mainly by means of deployment in the private chain
  • the smart contract in the ACC completes decentralized transaction information release and automatic configuration; in the alliance chain, the transaction information between members is mainly connectivity test information or data dependent transmission information, which is mainly completed with the help of smart contracts deployed in the alliance chain Decentralized transaction information release and automatic configuration.
  • each independent data center network selects a search strategy according to actual needs: depth-first traversal strategy, breadth-first traversal first-in-first-out strategy, breadth-first traversal first-in-last-out strategy, and breadth-first traversal priority queue strategy .
  • the mutual fault diagnosis information is formed into a transaction and released in the private chain, including the following steps:
  • Step 3.1 Each independent data center network randomly assigns a certain original node x, constructs its neighbor list N(x) according to the determined search strategy, and conducts fault mutual diagnosis test with neighbor node y in the neighbor list in turn.
  • the results of the diagnostic test are classified as and Four situations, where T means normal, F means fault, and T/F means unable to determine;
  • Step 3.2 The fault diagnosis result is used as the content of the transaction.
  • X and y are the initiators of the transaction at the same time. They respectively use the pseudonyms related to their public key information, and use their private keys to complete the digital signature of the transaction, and then publish the transaction on In the private chain, all members will receive the transaction information;
  • Step 3.3 The consensus node in the private chain finally records the transaction of the fault diagnosis result in the generated block through a certain consensus algorithm
  • Step 3.4 The member server that has completed the fault diagnosis will be marked. According to the search strategy, select another member server that has not been diagnosed or has not yet been diagnosed. Repeat steps 3.1-3.3 until all member servers in the independent data center network have been completed The fault is diagnosed or marked.
  • the smart contract is based on a certain diagnostic system model and diagnostic technology, and the faulty server is placed in the fault set, and the normal server is placed in the normal set.
  • smart contracts are divided into the following two situations: Optimistic diagnosis: Combining the diagnosis results of all neighbor servers in the data center network to determine and investigate one by one, and each server is classified into the fault set and the normal set. It is accurate; pessimistic diagnosis: the number of times the diagnosis result T of each server is more than the number of F is regarded as a normal server, and the number of times the diagnosis result F of each server is more than the number of T is regarded as a faulty server. If the diagnosis result of the server is only T/F, it is considered as a faulty server.
  • step 3 the processing of the member servers in the faulty set is completed through the smart contract, the backup server is used to replace the original faulty server or the tasks of the original faulty server are migrated to other normal servers, and the processing announcement is issued to the data center network.
  • recording transactions between independent data center networks that may be dependent on data transmission in the alliance chain and completing data transmission includes the following steps:
  • Step 4.2 Test the connectivity between data center network A and data center network B; if connected, data center network A initiates a transaction, uses the transmitted data feature as the transaction content, uses the pseudonym related to its public key information, and uses its private The key completes the digital signature of the transaction and sends it to data center network B; if it is not connected, submit a fault report to the smart contract;
  • Step 4.3 The consensus node in the alliance chain finally records the transaction in the generated block through a certain consensus algorithm
  • Step 4.4 Data center network A and B complete data transmission.
  • the published transaction content connectivity test information or data dependent transmission information can be verified and tracked through the Merkle tree and the block chain respectively; for the private chain, the published transaction content failure mutual diagnosis Information, server replacement information, or task adjustment information can be verified and tracked through the Merkle tree and the blocks of the blockchain, respectively.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Hardware Design (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Debugging And Monitoring (AREA)

Abstract

本发明提供一种基于混合链的数据中心网络故障诊断和自动配置方法，使用智能合约对超大规模的数据中心网络群组实现去中心化的故障诊断和系统自动配置。基于混合链完成超大规模数据中心网络架构建立，完成混合链中私有链和联盟链的安全参数初始化。在私有链中，利用智能合约完成独立数据中心网络的故障节点诊断，将识别出的故障节点放入故障集，将正常的节点放入正常集，同时发布交易并完成自动配置; 在联盟链中，实现有数据依赖关系的数据中心网络之间的交易发布和数据传输。对于任何交易均可以实现完整性验证和证据溯源。能保障以服务器为中心的新型的数据中心网络的去中心化、可靠性、可追踪性、鲁棒性和隐私性。

Description

一种基于混合链的数据中心网络故障诊断和自动配置方法 技术领域
本发明涉及主机安全和区块链的交叉技术领域，尤其涉及一种基于混合链的数据中心网络故障诊断和自动配置方法。
背景技术
随着云计算和云存储技术的发展，对于大容量数据传输网络的性能需求也日益增长，超大规模的数据中心网络应运而生。全球知名的IT企业，例如谷歌、百度、阿里、Facebook等，其数据存储和传输业务都高度依赖于数据中心网络。目前，全球以数据为中心的网络呈现指数级快速增长。为了降低成本，传统的以核心交换机为主导的数据中心网络结构，例如Fat-tree，正面临着严峻的挑战。研究者们致力于设计以服务器为中心的新型数据中心网络结构。该类型的数据中心网络，与传统的以核心交换机为主导的网络结构相比，可以有效的降低成本并提高路由的容错性，同时还具备大容量传输、传输口径小、传输效率高等优点。为了消除数据孤岛，使得多任务的业务处理流程更加简洁化和通畅化，存在数据相互依赖的各个数据中心网络之间不再各自为阵，而是形成了联盟群组。例如某大批量商品的数据，可能从生产环节的数据中心网络A迁移至销售环节的数据中心网络B。这种数据中心网络群组对于各个数据中心网络数据流的集中式管理提出了巨大的挑战。
此外，为了保障数据传输的可靠性，重要的途径之一就是排查数据中心网络的故障节点以提升网络的自动容错能力，以便于数据能够顺利的从发送端传输到接收端。因而，新型的以服务器为中心的数据中心网络(群组)结构，例如BCube、DCell、BCCC、RCube和GBC3等，都应当具备精准故障识别以及系统自动重配置等两大功能：前者的作用是将数据中心网络中的故障节点尽可能实时的发现并精准定位，后者的作用是通知数据中心网络的系统管理者将备用服务器替换掉故障服务器或者将故障服务器的数据任务分派给其它服务器，并告知其它有数据依赖关系的相邻的数据中心网络。在数据中心网络的自动故障诊断中，前者通常使用系统级诊断模型(例如PMC模型)结合t诊断或者t/t诊断技术来实现。后者可以通过故障日志结合事先约定的故障处理方法来实现。然而，这些传统的处理技术需要有中心化的处理机制，在故障记录、可靠性以及证据追溯等处理效果方面不够完善，尤其针对于超大规模的数据中心网络群组。
发明内容
本发明的目的是提供一种基于混合链的数据中心网络故障诊断和自动配置方法，将 超大规模的数据中心网络群组看成是具有数据依赖关系的联盟链，而每个独立的数据中心网络则是联盟链中的成员节点，同时它又作为独立的私有链，其中服务器或者普通交换设备作为私有链中的成员节点，从而形成了数据中心网络群组的链中链。
本发明提供一种基于混合链的数据中心网络故障诊断和自动配置方法，所述方法包括以下步骤：
步骤1：基于混合链的数据中心网络架构建立：将独立的数据中心网络架构设计成私有链，数据中心网络中的服务器作为私有链的成员，每个成员初始化其邻居列表；各个具有数据依赖关系的相邻的数据中心网络作为成员组成联盟链，同时形成数据中心网络群组，每个成员初始化其邻居列表；若干私有链和联盟链共同组成了混合区块链；
步骤2：区块链的安全参数初始化：独立的数据中心网络作为私有链，基于某种公钥密码体制，各成员获得其公私密钥对；私有链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到私有链区块中并可对交易信息的完整性进行验证；各成员通过假名地址进行交易；各个具有数据依赖关系的相邻的数据中心网络形成联盟链，基于某种公钥密码体制，各成员获得其在联盟链中的公私密钥对；联盟链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到联盟链区块中并可对交易信息的完整性进行验证；各成员同样通过假名地址进行交易；
步骤3：独立的数据中心网络完成故障节点诊断和系统自动配置：首先确定搜索策略，遍历整个数据中心网络成员；其次完成和邻居成员的故障互诊断测试，将故障互诊断信息形成交易发布在私有链中；然后，根据私有链中的交易信息，基于某种诊断系统模型和诊断技术，将故障成员放入故障集合，将正常成员放入正常集合；最终，通过智能合约完成对故障集合中成员服务器的处理并发布处理交易公告；
步骤4：联盟链中各个独立的数据中心网络之间可能存在数据传输依赖，两个相邻的独立数据中心网络通过各自的数据属性集合判定当前传输的数据是否存在数据依赖关系，如果存在数据依赖关系，则记录交易并完成数据传输；
步骤5：对数据中心网络群组已发布的所有交易记录实现验证和追踪。
进一步改进在于：所述步骤1中混合链中的私有链和联盟链的架构建立：在私有链中，成员间的交易信息主要为故障互诊断信息、服务器更换信息或者任务调整信息，主要借助于部署在私有链中的智能合约完成去中心化的交易信息发布和自动配置；在联盟链中，成员间的交易信息主要为连通性测试信息或者数据依赖传输信息，主要借助于部署在联盟链中的智能合约完成去中心化的交易信息发布和自动配置。
进一步改进在于：所述步骤3中，各独立的数据中心网络根据实际需求选择搜索策略：深度优先遍历策略、广度优先遍历的先进先出策略、广度优先遍历的先进后出策略和广度优先遍历的优先权队列策略。
进一步改进在于：所述步骤3中，在完成和邻居成员的故障互诊断测试时，将故障互诊断信息形成交易发布在私有链中，包括如下步骤：
步骤3.1：各独立的数据中心网络随机指定某个原始节点x，按照确定的搜索策略，构建其邻居列表N(x)并依次和邻居列表中的邻居节点y，进行故障互诊断测试，故障互诊断测试的结果分类为
Figure PCTCN2021073501-appb-000001
和
Figure PCTCN2021073501-appb-000002
四种情况，其中T表示正常，F表示故障，T/F表示无法确定；
步骤3.2：故障诊断结果作为交易的内容，x和y同时作为交易的发起者，分别使用和其公钥信息相关的假名，利用其私钥分别完成对交易的数字签名，然后将该交易发布在私有链中，所有成员都将收到该交易信息；
步骤3.3：私有链中的共识节点通过某种共识算法最终将该故障诊断结果的交易记录在产生的区块中；
步骤3.4：完成故障诊断的成员服务器将被标记，按照搜索策略选择另一个未进行诊断或者还未诊断出结果的成员服务器重复步骤3.1-3.3，直至独立数据中心网络中的所有成员服务器均已完成故障诊断或者被标记。
进一步改进在于：所述步骤3中根据私有链中记录的交易信息，智能合约基于某种诊断系统模型和诊断技术，将故障服务器放入故障集合，将正常服务器放入正常集合。智能合约根据不同的诊断模型和诊断技术，划分为如下两种情况：乐观诊断：结合数据中心网络中所有邻居服务器的诊断结果进行逐一判定和排查，每一服务器划入故障集合和正常集合的结果是精确的；悲观诊断：对每一服务器的诊断结果T的次数多于F的次数则认为是正常服务器，对每一服务器的诊断结果F的次数多于T的次数则认为是故障服务器，若服务器的诊断结果只有T/F则认为是故障服务器。
进一步改进在于：所述步骤3中，通过智能合约完成对故障集合中成员服务器的处理，使用备份服务器替换原故障服务器或者将原故障服务器的任务迁移至其它正常的服务器，并发布处理公告给数据中心网络。
进一步改进在于：所述步骤4中联盟链中可能存在数据传输依赖的各个独立数据中心网络之间记录交易并完成数据传输包括以下步骤：
步骤4.1：各独立的数据中心网络，标记为A，作为联盟链成员，对当前传输的数据根据其 特征构建属性集T A＝{a 1,a 2,…,a n}，其相邻的数据中心网络，标记为B若存在属性集T B＝{b 1,b 2,…,b m}，且T AΔT B！＝Ф则与A有数据依赖关系；
步骤4.2：数据中心网络A测试与数据中心网络B之间的连通性；若连通，数据中心网络A发起交易，将传输数据特征作为交易内容，使用和其公钥信息相关的假名，利用其私钥完成对交易的数字签名，发送给数据中心网络B；若不连通，则提交故障申告给智能合约；
步骤4.3：联盟链中的共识节点通过某种共识算法最终将该交易记录在产生的区块中；
步骤4.4：数据中心网络A和B完成数据传输。
进一步改进在于：所述步骤五中对于联盟链，发布的交易内容连通性测试信息或者数据依赖传输信息可以通过Merkle树和区块链的区块分别进行验证和追踪；对于私有链，发布的交易内容故障互诊断信息、服务器更换信息或者任务调整信息可以通过Merkle树和区块链的区块分别进行验证和追踪。
本发明有益效果：将区块链技术应用于以服务器为中心的新型的数据中心网络结构中，利用智能合约实现了去中心化的管理和系统的自动配置；基于混合区块链结构，独立的数据中心网络内部采用私有链结构，非授权内部节点不能加入私有链，故障诊断结果均以交易形式记录在私有链中，各个可能存在数据依赖的独立中心网络组成联盟链，对所有的数据流向均要记录在联盟链中，实现了不可篡改和可追溯的功能，且采用了Merkle树机制对所有交易的完整性进行验证，有效解决了故障记录的可靠性问题；使用区块链技术增强了超大规模数据中心网络的可靠性和数据传输的鲁棒性，假名的使用也增加了匿名性和隐私保护。
附图说明
图1是本发明的数据中心网络群组架构图。
图2是本发明的区块链的区块结构图。
图3是本发明的步骤执行流程图。
具体实施方式
为了加深对本发明的理解，下面将结合实施例对本发明作进一步详述，该实施例仅用于解释本发明，并不构成对本发明保护范围的限定。
如图1-3所示，本实施例提供一种基于混合链的数据中心网络故障诊断和自动配置方法，所述方法包括以下步骤：
步骤1：基于混合链的数据中心网络架构建立：将独立的数据中心网络架构设计成私有链，数据中心网络中的服务器作为私有链的成员，每个成员初始化其邻居列表；各个具有数据依赖关系的相邻的数据中心网络作为成员组成联盟链，同时形成数据中心网络群组，每个成员 初始化其邻居列表；若干私有链和联盟链共同组成了混合区块链；
步骤2：区块链的安全参数初始化：独立的数据中心网络作为私有链，基于某种公钥密码体制，各成员获得其公私密钥对；私有链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到私有链区块中并可对交易信息的完整性进行验证；各成员通过假名地址进行交易；各个具有数据依赖关系的相邻的数据中心网络形成联盟链，基于某种公钥密码体制，各成员获得其在联盟链中的公私密钥对；联盟链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到联盟链区块中并可对交易信息的完整性进行验证；各成员同样通过假名地址进行交易；
步骤3：独立的数据中心网络完成故障节点诊断和系统自动配置：首先确定搜索策略，遍历整个数据中心网络成员；其次完成和邻居成员的故障互诊断测试，将故障互诊断信息形成交易发布在私有链中；然后，根据私有链中的交易信息，基于某种诊断系统模型和诊断技术，将故障成员放入故障集合，将正常成员放入正常集合；最终，通过智能合约完成对故障集合中成员服务器的处理并发布处理交易公告；
步骤4：联盟链中各个独立的数据中心网络之间可能存在数据传输依赖，两个相邻的独立数据中心网络通过各自的数据属性集合判定当前传输的数据是否存在数据依赖关系，如果存在数据依赖关系，则记录交易并完成数据传输；
步骤5：对数据中心网络群组已发布的所有交易记录实现验证和追踪。
所述步骤1中混合链中的私有链和联盟链的架构建立：在私有链中，成员间的交易信息主要为故障互诊断信息、服务器更换信息或者任务调整信息，主要借助于部署在私有链中的智能合约完成去中心化的交易信息发布和自动配置；在联盟链中，成员间的交易信息主要为连通性测试信息或者数据依赖传输信息，主要借助于部署在联盟链中的智能合约完成去中心化的交易信息发布和自动配置。
所述步骤3中，各独立的数据中心网络根据实际需求选择搜索策略：深度优先遍历策略、广度优先遍历的先进先出策略、广度优先遍历的先进后出策略和广度优先遍历的优先权队列策略。
所述步骤3中，在完成和邻居成员的故障互诊断测试时，将故障互诊断信息形成交易发布在私有链中，包括如下步骤：
步骤3.1：各独立的数据中心网络随机指定某个原始节点x，按照确定的搜索策略，构建其邻居列表N(x)并依次和邻居列表中的邻居节点y，进行故障互诊断测试，故障互诊断测试的结果分类为
Figure PCTCN2021073501-appb-000003
和
Figure PCTCN2021073501-appb-000004
四 种情况，其中T表示正常，F表示故障，T/F表示无法确定；
步骤3.2：故障诊断结果作为交易的内容，x和y同时作为交易的发起者，分别使用和其公钥信息相关的假名，利用其私钥分别完成对交易的数字签名，然后将该交易发布在私有链中，所有成员都将收到该交易信息；
步骤3.3：私有链中的共识节点通过某种共识算法最终将该故障诊断结果的交易记录在产生的区块中；
步骤3.4：完成故障诊断的成员服务器将被标记，按照搜索策略选择另一个未进行诊断或者还未诊断出结果的成员服务器重复步骤3.1-3.3，直至独立数据中心网络中的所有成员服务器均已完成故障诊断或者被标记。
所述步骤3中根据私有链中记录的交易信息，智能合约基于某种诊断系统模型和诊断技术，将故障服务器放入故障集合，将正常服务器放入正常集合。智能合约根据不同的诊断模型和诊断技术，划分为如下两种情况：乐观诊断：结合数据中心网络中所有邻居服务器的诊断结果进行逐一判定和排查，每一服务器划入故障集合和正常集合的结果是精确的；悲观诊断：对每一服务器的诊断结果T的次数多于F的次数则认为是正常服务器，对每一服务器的诊断结果F的次数多于T的次数则认为是故障服务器，若服务器的诊断结果只有T/F则认为是故障服务器。
所述步骤3中，通过智能合约完成对故障集合中成员服务器的处理，使用备份服务器替换原故障服务器或者将原故障服务器的任务迁移至其它正常的服务器，并发布处理公告给数据中心网络。
所述步骤4中联盟链中可能存在数据传输依赖的各个独立数据中心网络之间记录交易并完成数据传输包括以下步骤：
步骤4.1：各独立的数据中心网络，标记为A，作为联盟链成员，对当前传输的数据根据其特征构建属性集T A＝{a 1,a 2,…,a n}，其相邻的数据中心网络，标记为B若存在属性集T B＝{b 1,b 2,…,b m}，且T AΔT B！＝Ф则与A有数据依赖关系；
步骤4.2：数据中心网络A测试与数据中心网络B之间的连通性；若连通，数据中心网络A发起交易，将传输数据特征作为交易内容，使用和其公钥信息相关的假名，利用其私钥完成对交易的数字签名，发送给数据中心网络B；若不连通，则提交故障申告给智能合约；
步骤4.3：联盟链中的共识节点通过某种共识算法最终将该交易记录在产生的区块中；
步骤4.4：数据中心网络A和B完成数据传输。
所述步骤五中对于联盟链，发布的交易内容连通性测试信息或者数据依赖传输信息 可以通过Merkle树和区块链的区块分别进行验证和追踪；对于私有链，发布的交易内容故障互诊断信息、服务器更换信息或者任务调整信息可以通过Merkle树和区块链的区块分别进行验证和追踪。

Claims (9)

  1. 一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述方法包括以下步骤：
    步骤1：基于混合链的数据中心网络架构建立：将独立的数据中心网络架构设计成私有链，数据中心网络中的服务器作为私有链的成员，每个成员初始化其邻居列表；各个具有数据依赖关系的相邻的数据中心网络作为成员组成联盟链，同时形成数据中心网络群组，每个成员初始化其邻居列表；若干私有链和联盟链共同组成了混合区块链；
    步骤2：区块链的安全参数初始化：独立的数据中心网络作为私有链，基于某种公钥密码体制，各成员获得其公私密钥对；私有链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到私有链区块中并可对交易信息的完整性进行验证；各成员通过假名地址进行交易；各个具有数据依赖关系的相邻的数据中心网络形成联盟链，基于某种公钥密码体制，各成员获得其在联盟链中的公私密钥对；联盟链采用Merkle树形式存储各成员的交易信息哈希值路径；对Merkle树的树根签名后保存到联盟链区块中并可对交易信息的完整性进行验证；各成员同样通过假名地址进行交易；
    步骤3：独立的数据中心网络完成故障节点诊断和系统自动配置：首先确定搜索策略，遍历整个数据中心网络成员；其次完成和邻居成员的故障互诊断测试，将故障互诊断信息形成交易发布在私有链中；然后，根据私有链中的交易信息，基于某种诊断系统模型和诊断技术，将故障成员放入故障集合，将正常成员放入正常集合；最终，通过智能合约完成对故障集合中成员服务器的处理并发布处理交易公告；
    步骤4：联盟链中各个独立的数据中心网络之间可能存在数据传输依赖，两个相邻的独立数据中心网络通过各自的数据属性集合判定当前传输的数据是否存在数据依赖关系，如果存在数据依赖关系，则记录交易并完成数据传输；
    步骤5：对数据中心网络群组已发布的所有交易记录实现验证和追踪。
  2. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤1中混合链中的私有链和联盟链的架构建立：在私有链中，成员间的交易信息主要为故障互诊断信息、服务器更换信息或者任务调整信息，主要借助于部署在私有链中的智能合约完成去中心化的交易信息发布和自动配置；在联盟链中，成员间的交易信息主要为连通性测试信息或者数据依赖传输信息，主要借助于部署在联盟链中的智能合约完成去中心化的交易信息发布和自动配置。
  3. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤3中，各独立的数据中心网络根据实际需求选择搜索策略：深度优先遍历策 略、广度优先遍历的先进先出策略、广度优先遍历的先进后出策略和广度优先遍历的优先权队列策略。
  4. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤3中，在完成和邻居成员的故障互诊断测试时，将故障互诊断信息形成交易发布在私有链中，包括如下步骤：
    步骤3.1：各独立的数据中心网络随机指定某个原始节点x，按照确定的搜索策略，构建其邻居列表N(x)并依次和邻居列表中的邻居节点y，进行故障互诊断测试，故障互诊断测试的结果分类为
    Figure PCTCN2021073501-appb-100001
    和
    Figure PCTCN2021073501-appb-100002
    四种情况，其中T表示正常，F表示故障，T/F表示无法确定；
    步骤3.2：故障诊断结果作为交易的内容，x和y同时作为交易的发起者，分别使用和其公钥信息相关的假名，利用其私钥分别完成对交易的数字签名，然后将该交易发布在私有链中，所有成员都将收到该交易信息；
    步骤3.3：私有链中的共识节点通过某种共识算法最终将该故障诊断结果的交易记录在产生的区块中；
    步骤3.4：完成故障诊断的成员服务器将被标记，按照搜索策略选择另一个未进行诊断或者还未诊断出结果的成员服务器重复步骤3.1-3.3，直至独立数据中心网络中的所有成员服务器均已完成故障诊断或者被标记。
  5. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤3中根据私有链中记录的交易信息，智能合约基于某种诊断系统模型和诊断技术，将故障服务器放入故障集合，将正常服务器放入正常集合。
  6. 智能合约根据不同的诊断模型和诊断技术，划分为如下两种情况：乐观诊断：结合数据中心网络中所有邻居服务器的诊断结果进行逐一判定和排查，每一服务器划入故障集合和正常集合的结果是精确的；悲观诊断：对每一服务器的诊断结果T的次数多于F的次数则认为是正常服务器，对每一服务器的诊断结果F的次数多于T的次数则认为是故障服务器，若服务器的诊断结果只有T/F则认为是故障服务器。
  7. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤3中，通过智能合约完成对故障集合中成员服务器的处理，使用备份服务器替换原故障服务器或者将原故障服务器的任务迁移至其它正常的服务器，并发布处理公告给数据中心网络。
  8. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征 在于：所述步骤4中联盟链中可能存在数据传输依赖的各个独立数据中心网络之间记录交易并完成数据传输包括以下步骤：
    步骤4.1：各独立的数据中心网络，标记为A，作为联盟链成员，对当前传输的数据根据其特征构建属性集T A＝{a 1,a 2,…,a n}，其相邻的数据中心网络，标记为B若存在属性集T B＝{b 1,b 2,…,b m}，且T A∧T B！＝Ф则与A有数据依赖关系；
    步骤4.2：数据中心网络A测试与数据中心网络B之间的连通性；若连通，数据中心网络A发起交易，将传输数据特征作为交易内容，使用和其公钥信息相关的假名，利用其私钥完成对交易的数字签名，发送给数据中心网络B；若不连通，则提交故障申告给智能合约；
    步骤4.3：联盟链中的共识节点通过某种共识算法最终将该交易记录在产生的区块中；
    步骤4.4：数据中心网络A和B完成数据传输。
  9. 如权利要求1所述的一种基于混合链的数据中心网络故障诊断和自动配置方法，其特征在于：所述步骤5中对于联盟链，发布的交易内容连通性测试信息或者数据依赖传输信息可以通过Merkle树和区块链的区块分别进行验证和追踪；对于私有链，发布的交易内容故障互诊断信息、服务器更换信息或者任务调整信息可以通过Merkle树和区块链的区块分别进行验证和追踪。
PCT/CN2021/073501 2020-06-08 2021-01-25 一种基于混合链的数据中心网络故障诊断和自动配置方法 WO2021248917A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010511243.7 2020-06-08
CN202010511243.7A CN111405074B (zh) 2020-06-08 2020-06-08 一种基于混合链的数据中心网络故障诊断和自动配置方法

Publications (1)

Publication Number Publication Date
WO2021248917A1 true WO2021248917A1 (zh) 2021-12-16

Family

ID=71433720

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/073501 WO2021248917A1 (zh) 2020-06-08 2021-01-25 一种基于混合链的数据中心网络故障诊断和自动配置方法

Country Status (2)

Country Link
CN (1) CN111405074B (zh)
WO (1) WO2021248917A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114928558A (zh) * 2022-06-14 2022-08-19 上海万向区块链股份公司 基于区块链的运维方法和系统
CN115933565A (zh) * 2022-12-23 2023-04-07 广东职业技术学院 一种agv任务交换方法、装置、系统和介质

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111405074B (zh) * 2020-06-08 2020-09-08 南京邮电大学 一种基于混合链的数据中心网络故障诊断和自动配置方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108769150A (zh) * 2018-05-14 2018-11-06 百度在线网络技术（北京）有限公司 区块链网络的数据处理方法、装置、集群节点和存储介质
US10158527B2 (en) * 2016-10-28 2018-12-18 International Business Machines Corporation Changing an existing blockchain trust configuration
CN110113185A (zh) * 2019-04-18 2019-08-09 卓尔购电子商务(武汉)有限公司 一种快速部署联盟链的方法、系统、存储介质和装置
CN110380893A (zh) * 2019-06-19 2019-10-25 深圳壹账通智能科技有限公司 通讯故障分析方法、装置及区块链通讯平台
CN110569909A (zh) * 2019-09-10 2019-12-13 腾讯科技（深圳）有限公司 基于区块链的故障预警方法、装置、设备及存储介质
CN111405074A (zh) * 2020-06-08 2020-07-10 南京邮电大学 一种基于混合链的数据中心网络故障诊断和自动配置方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102170648A (zh) * 2011-01-28 2011-08-31 北京浩阳华夏科技有限公司 无线传感器网络的被动诊断方法
WO2018161007A1 (en) * 2017-03-03 2018-09-07 Mastercard International Incorporated Method and system for storage and transfer of verified data via blockhain
CN109496407B (zh) * 2018-07-27 2020-07-24 袁振南 区块链系统中的消息传输方法、装置及存储介质
CN109033405B (zh) * 2018-08-03 2020-09-08 华为技术有限公司 维护区块链的方法和装置、服务器和计算机可读存储介质
CN109361565A (zh) * 2018-11-08 2019-02-19 国网黑龙江省电力有限公司信息通信公司 基于区块链的数据中心监测管理系统
CN110661658B (zh) * 2019-09-23 2022-08-12 腾讯科技（深圳）有限公司 一种区块链网络的节点管理方法、装置及计算机存储介质
CN110781200B (zh) * 2019-10-30 2022-07-08 百度在线网络技术（北京）有限公司 一种区块链异常数据的处理方法、装置、设备和介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10158527B2 (en) * 2016-10-28 2018-12-18 International Business Machines Corporation Changing an existing blockchain trust configuration
CN108769150A (zh) * 2018-05-14 2018-11-06 百度在线网络技术（北京）有限公司 区块链网络的数据处理方法、装置、集群节点和存储介质
CN110113185A (zh) * 2019-04-18 2019-08-09 卓尔购电子商务(武汉)有限公司 一种快速部署联盟链的方法、系统、存储介质和装置
CN110380893A (zh) * 2019-06-19 2019-10-25 深圳壹账通智能科技有限公司 通讯故障分析方法、装置及区块链通讯平台
CN110569909A (zh) * 2019-09-10 2019-12-13 腾讯科技（深圳）有限公司 基于区块链的故障预警方法、装置、设备及存储介质
CN111405074A (zh) * 2020-06-08 2020-07-10 南京邮电大学 一种基于混合链的数据中心网络故障诊断和自动配置方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114928558A (zh) * 2022-06-14 2022-08-19 上海万向区块链股份公司 基于区块链的运维方法和系统
CN114928558B (zh) * 2022-06-14 2023-12-12 上海万向区块链股份公司 基于区块链的运维方法和系统
CN115933565A (zh) * 2022-12-23 2023-04-07 广东职业技术学院 一种agv任务交换方法、装置、系统和介质
CN115933565B (zh) * 2022-12-23 2023-10-20 广东职业技术学院 一种agv任务交换方法、装置、系统和介质

Also Published As

Publication number Publication date
CN111405074A (zh) 2020-07-10
CN111405074B (zh) 2020-09-08

Similar Documents

Publication Publication Date Title
WO2021248917A1 (zh) 一种基于混合链的数据中心网络故障诊断和自动配置方法
US11797489B2 (en) System and method for using local storage to emulate centralized storage
CN107193490B (zh) 一种基于区块链的分布式数据存储系统及方法
US11294888B2 (en) Blockchain system and blockchain transaction data processing method based on ethereum
US20190129895A1 (en) Distributed blockchain oracle
US20180308091A1 (en) Fairness preserving byzantine agreements
KR20210133289A (ko) 블록체인 네트워크에서 데이터 추출
CN111612455A (zh) 一种面向用电信息保护的拜占庭容错联盟链共识方法及其系统、存储介质
US20110022883A1 (en) Method for Voting with Secret Shares in a Distributed System
CN111144883B (zh) 一种区块链网络的处理性能分析方法及装置
US20070016822A1 (en) Policy-based, cluster-application-defined quorum with generic support interface for cluster managers in a shared storage environment
WO2020108289A1 (zh) 一种数据库系统、节点和方法
US9983823B1 (en) Pre-forking replicas for efficient scaling of a distribued data storage system
Wang et al. Improvement research of PBFT consensus algorithm based on credit
US11816069B2 (en) Data deduplication in blockchain platforms
Manevich et al. Endorsement in Hyperledger Fabric via service discovery
US9935844B2 (en) Reducing internodal communications in a clustered system
Tian et al. A byzantine fault-tolerant raft algorithm combined with Schnorr signature
CN111241590A (zh) 一种数据库系统、节点和方法
WO2023050986A1 (zh) 维护区块链系统的网络架构信息
Gupta et al. Reliable transactions in serverless-edge architecture
CN103399943A (zh) 集群数据库并行查询的通讯方法和通讯装置
EP3349416B1 (en) Relationship chain processing method and system, and storage medium
WO2023179056A1 (zh) 区块链网络的共识处理方法、装置、设备及存储介质、程序产品
WO2021201827A1 (en) Method and apparatus maintaining private data with consortium blockchain

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: 21821172

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: 21821172

Country of ref document: EP

Kind code of ref document: A1