WO2015109804A1 - 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法 - Google Patents

一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法 Download PDF

Info

Publication number
WO2015109804A1
WO2015109804A1 PCT/CN2014/083113 CN2014083113W WO2015109804A1 WO 2015109804 A1 WO2015109804 A1 WO 2015109804A1 CN 2014083113 W CN2014083113 W CN 2014083113W WO 2015109804 A1 WO2015109804 A1 WO 2015109804A1
Authority
WO
WIPO (PCT)
Prior art keywords
backup
virtual machine
primary
server
dual
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2014/083113
Other languages
English (en)
French (fr)
Inventor
管海兵
马汝辉
李健
戚正伟
钱正宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Jiao Tong University
Original Assignee
Shanghai Jiao Tong University
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 Shanghai Jiao Tong University filed Critical Shanghai Jiao Tong University
Priority to US14/412,125 priority Critical patent/US20160323427A1/en
Publication of WO2015109804A1 publication Critical patent/WO2015109804A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/40—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/202—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant
    • G06F11/2038—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant with a single idle spare processing component
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/202—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant
    • G06F11/2048—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant where the redundant components share neither address space nor persistent storage
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/2097—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements maintaining the standby controller/processing unit updated
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • 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
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45595—Network integration; Enabling network access in virtual machine instances
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2201/00—Indexing scheme relating to error detection, to error correction, and to monitoring
    • G06F2201/815—Virtual
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
    • H04L43/0817—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/20—Arrangements for monitoring or testing data switching networks the monitoring system or the monitored elements being virtualised, abstracted or software-defined entities, e.g. SDN or NFV

Definitions

  • the present invention relates to a highly reliable disaster recovery technology in a virtualized environment, and in particular, to a Dual-system hot backup disaster recovery system and method for network services in a virtualized environment.
  • Checkpointing The technology uses the two physical devices to form the primary server/backup server mode, and backs up the same application/virtual machine.
  • the virtual machine migration technology periodically backs up the status of the primary server virtual machine to the backup server to implement disaster recovery.
  • the virtual machine of the standby server is in a non-operation state. After the primary server fails, it can quickly restore to the previous state of the primary server, and continue to retain all the original network connections, so that the client does not feel that the server has failed and is faulty. Recovery.
  • periodic frequent backups (20-40ms once) must be performed, resulting in greatly reduced throughput of the primary server and excessive CPU overhead.
  • Checkpointing technology saves all the packets sent by the server to the client in a buffer. Only when one backup is completed can these packets be released, which leads to an increase in network latency.
  • Lockstepping The technology adopts the parallel running mode of the dual-system to ensure that the status of the backup server of the primary server is consistent, so that after the primary server fails, the client can directly connect with the backup server to quickly recover from the fault.
  • Lockstepping technology can only be applied to the case of assigning a single processor to a virtual machine.
  • the performance scalability of a multiprocessor virtual machine is very poor, and the performance of a virtual machine more than a dual processor is reduced to 1/7 of that of a single processor.
  • the virtual machines of the primary backup server can run directly in parallel, and for non-determined instructions, instruction level synchronization needs to be performed between the primary backup server virtual machines, which also increases the overhead of the system.
  • the present invention provides a dual-system hot backup disaster tolerance system.
  • the primary virtual machine and the backup virtual machine run in parallel, and the respective output results are generated according to the request sent by the client, and the output results of the primary virtual machine and the backup virtual machine are compared. If they are inconsistent, the backup needs to be performed, thus ensuring the failure.
  • the invention provides A dual-system hot backup disaster recovery system is used for network services in a virtualized environment.
  • the dual-system hot backup disaster recovery system includes a primary server and a backup server, and the primary server and the backup server are connected through a network, and are characterized in that: the primary server Running the primary virtual machine, running the backup virtual machine on the backup server, the backup virtual machine is in the application layer semantic alternative state of the primary virtual machine, and the application layer semantic alternative state means that the backup virtual machine can replace the primary virtual in the application layer semantics.
  • the machine performs the service and produces the correct output for any client request.
  • the primary server sends the client request to the primary virtual machine and the backup virtual machine, and the primary virtual machine and the backup virtual machine run in parallel to generate respective response data packets.
  • the dual-system hot backup disaster recovery system further includes a primary backup manager running on the primary virtual machine, and a backup backup manager running on the backup virtual machine, and the backup backup manager is configured to generate response data generated by the backup virtual machine.
  • the package is sent to the primary backup manager.
  • the primary backup manager is used to compare whether the response packets of the primary virtual machine and the standby virtual machine are consistent. If the backup virtual machine is in an alternate state of the primary virtual machine, the primary backup manager will be the primary virtual machine.
  • the machine-generated response packet is sent to the client; if it is inconsistent, the standby virtual machine is not in the alternative state of the primary virtual machine. .
  • the primary backup manager will present the current virtual machine Back up to the standby virtual machine.
  • the backup is a non-periodic backup.
  • the backup to the standby virtual machine is an incremental backup.
  • Incremental backup is used in the system to reduce the overhead of state backup.
  • the dual-machine parallel operation in the present invention so the state of the backup virtual machine also changes between the two state backups, which makes it unnecessary to back up only the primary virtual machine state increment.
  • the method of space-for-time is employed in the present invention.
  • the backup backup manager detects the main virtual The heartbeat packet of the virtual machine, if the backup backup manager does not receive the heartbeat packet of the primary virtual machine, the client requests the data packet to directly reach the backup virtual machine, and after the backup virtual machine generates the response data packet, the backup backup manager will The response packet is sent directly to the client.
  • a heartbeat packet mechanism is introduced in the system to monitor whether the primary virtual machine continues to survive. If the backup virtual machine does not receive the heartbeat packet, it considers that the primary virtual machine has failed and will take failover measures to replace the primary virtual machine to continue providing services. In this case, the request packet sent by the client will directly reach the backup virtual machine. After the backup virtual machine generates the response packet, it will not be sent to the primary virtual machine, but will be sent directly to the client. In this case, the source of the packet received by the client is changed from the primary virtual machine to the backup virtual machine, and the server does not find a fast failure recovery.
  • the shadow page table mechanism provided by the virtual machine monitor is enabled to get the page that was modified after the last state backup.
  • the basic principle is to change the pages of all virtual machines to write protection, so that once a page is written, an exception is triggered and the exception handler is entered.
  • the invention also provides a dual-system hot backup disaster recovery method, which comprises the following steps:
  • the primary server sends the request sent by the client to the primary virtual machine and the backup virtual machine respectively through flow control;
  • the backup backup manager sends the response packet generated by the backup virtual machine to the primary backup manager
  • the primary backup manager is used to compare the response packets of the primary virtual machine and the backup virtual machine. If the backup virtual machine is in the alternate state of the application layer semantics of the primary virtual machine, the response data packet of the primary virtual machine is sent to the client. end If the inconsistency, the standby virtual machine is not in the alternate state of the application layer semantics of the primary virtual machine, the primary backup manager backs up the current state of the primary virtual machine to the standby virtual machine.
  • the dual-system hot backup disaster tolerance system and the method thereof provided by the present invention have the following beneficial technical effects:
  • the system implementation solves the technical problems of consistency of storage access, consistency of network protocol, consistency of CPU instructions of multi-core state in the case of parallel connection of the primary backup server.
  • the backup of the primary server status in the solution is aperiodic, the backup interval is greater than 1 second, and the frequency is reduced by more than two orders of magnitude relative to the prior art, which greatly reduces system overhead and substantially eliminates virtual machine state. Backups interfere with the performance of the primary server.
  • the main server of the present invention can deliver the output result without waiting for the backup to be completed, thereby improving the throughput of the system.
  • the solution of the present invention can provide fast disaster recovery recovery, and the disaster recovery time for network services and database services is faster than the prior art.
  • FIG. 3 is a schematic flowchart of a dual-system hot backup disaster recovery system according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of a process of incremental backup of a dual-system hot backup disaster recovery system according to an embodiment of the present invention.
  • FIG. 1 is a schematic flow chart of an existing checkpoint technique.
  • the primary virtual machine processes the client request and generates a response, and the standby virtual machine is in a non-operational state.
  • the timing module generates periodic events.
  • the backup manager obtains the state of the primary virtual machine, and backs up the changed state after the last backup to the backup virtual machine.
  • FIG. 2 is a schematic flow chart of an existing step lock technique.
  • the primary virtual machine and the backup virtual machine execute the request sent by the client in parallel, and the primary virtual machine sends a response back to the client. Because of non-deterministic instructions (such as memory access, clock interrupts, etc.), you need to do instruction-level synchronization between virtual machines to avoid differences in state between the two sides.
  • the present invention provides a dual-system hot backup disaster recovery system for network services in a virtualized environment.
  • the dual-system hot backup disaster recovery system includes a primary server and a backup server, and the primary server and the backup server are connected through a network, and the features are:
  • the primary virtual machine runs on the primary server, and the backup virtual machine runs on the backup server.
  • the backup virtual machine is in the application layer semantic alternative state of the primary virtual machine.
  • the semantic alternative state of the application layer refers to the backup virtual machine in the application layer semantics. Instead of the primary virtual machine for service, it produces the correct output for any client request.
  • the request packet sent by the client first arrives at the peripheral switch, and the switch determines the forwarding port by the destination MAC address.
  • the virtual machine MAC address corresponding to the switch learns the port as the primary server NIC port, so the request packet is sent to the primary server.
  • the primary server sends the client request to the primary virtual machine and the backup virtual machine respectively, and the primary virtual machine and the backup virtual machine run in parallel to generate respective response data packets.
  • the dual-system hot backup disaster recovery system also includes a primary backup manager running on the primary virtual machine and a backup backup manager running on the backup virtual machine, and the backup backup manager is configured to send the response data packet generated by the backup virtual machine to
  • the primary backup manager is used to compare whether the response packets of the primary virtual machine and the backup virtual machine are consistent. If they are consistent, the backup virtual machine is in an alternative state of the primary virtual machine, and the primary backup manager generates the primary virtual machine.
  • the response packet is sent to the client; if it is inconsistent, the standby virtual machine is not in an alternate state of the primary virtual machine .
  • the primary backup manager backs up the current state of the primary virtual machine to the standby virtual machine.
  • Backup is a non-periodic backup.
  • Incremental backup is used in the system to reduce the overhead of state backup.
  • the dual-machine parallel operation in the present invention so the state of the backup virtual machine also changes between the two state backups, which makes it unnecessary to back up only the primary virtual machine state increment.
  • the method of space-for-time is employed in the present invention.
  • the heartbeat packet of the virtual machine if the backup backup manager does not receive the heartbeat packet of the primary virtual machine, the client requests the data packet to directly reach the backup virtual machine, and after the backup virtual machine generates the response data packet, the backup backup manager will The response packet is sent directly to the client.
  • a heartbeat packet mechanism is introduced in the system to monitor whether the primary virtual machine continues to survive. If the backup virtual machine does not receive the heartbeat packet, it considers that the primary virtual machine has failed and will take failover measures to replace the primary virtual machine to continue providing services.
  • the backup server will send an ARP packet to the switch whose source MAC address is the MAC address of the standby virtual machine. This allows the switch to learn a new MAC Address-to-port mapping entry. After that, the destination MAC address sent by the client is the virtual machine's data packet, which will be sent directly to the backup server's network card. After the backup virtual machine generates the response packet, it is no longer sent to the primary virtual machine, but is sent directly to the client. In this case, the source of the packet received by the client is changed from the primary virtual machine to the backup virtual machine, and the server does not find a fast failure recovery.
  • the shadow page table mechanism provided by the virtual machine monitor is enabled. Gets which pages have been modified since the last state backup.
  • the basic principle is to change the pages of all virtual machines to write protection, so that once a page is written, an exception is triggered and the exception handler is entered. With the help of the 'shadow page table' mechanism, it is easy to get which pages have been modified since the last state backup.
  • FIG. 3 is a schematic flowchart of a dual-system hot backup disaster recovery system according to the embodiment, and the specific process is as follows:
  • Step 1 The primary server distributes the request packet sent by the client to the primary virtual machine and the backup virtual machine.
  • the process is as follows: First, the request packet sent by the client is sent by the switch to the primary server through the peripheral switch.
  • the main server receives the data packet and sends it to the software bridge.
  • the Linux tool TC Traffic
  • Control to intercept and distribute network packets, and send the packets to the primary virtual machine and the backup virtual machine.
  • the configuration method of the TC is as follows:
  • Step 2 The primary virtual machine and the backup virtual machine execute in parallel according to the application layer semantics, and generate respective outputs, and the backup virtual machine sends the output to the primary server.
  • the TC is configured to implement interception and forwarding of the backup VM output.
  • Step 3 The manager of the primary server compares whether the primary virtual machine and the backup virtual machine generate their respective outputs to satisfy the alternative rule. Specifically, two virtual interfaces in the form of queues are implemented in the manager, and the outputs of the primary virtual machine and the backup virtual machine are respectively redirected into one interface. The manager compares the packets in the two queues one by one to determine whether the backup virtual machine is still an alternative state of the primary virtual machine. The TC is configured to redirect the output.
  • the specific method is as follows:
  • Step 4 Send the output of the primary server as a response packet to the client.
  • Step 5 If it is determined that the backup virtual machine is not an alternative state of the primary virtual machine, the current state of the primary virtual machine is backed up to the backup virtual machine.
  • FIG. 4 is a schematic diagram of a process of incremental backup of the dual-system hot backup disaster recovery system of the embodiment.
  • Step 1 The backup manager on the primary server obtains the state change part of the primary virtual machine after the last backup.
  • Step 2 The Backup Manager sends the changed part to the standby virtual machine.
  • Step 3 The backup virtual machine will update the partial cache temporarily.
  • Step 4 Back up all the temporary cache contents into the backup virtual machine.
  • the disk drive is interrupted by the primary virtual machine and the backup virtual machine by modifying the backend driver of the disk device.
  • the disk write data of the primary virtual machine and the standby virtual machine between the two backups is temporarily saved in their respective temporary caches.
  • the device status involves the front-end model of the virtual machine monitor, it is difficult to obtain. Therefore, the state before the primary virtual machine and the backup virtual machine device are discarded is selected. When the backup is complete, re-establish the connection to keep the device status consistent.
  • the dual-system hot backup disaster tolerance system and method thereof provided by the present invention, The technical problem of consistency of storage access, consistency of network protocol, consistency of CPU instructions of multi-core state, and the like in the case of parallel operation of the primary backup server is solved.
  • the backup of the status of the primary server in the solution is aperiodic. Sex, the backup interval is greater than 1 second, and the frequency is reduced by more than two orders of magnitude relative to the prior art, which greatly reduces the system overhead and basically eliminates the performance interference of the virtual machine state backup to the primary server; the primary server does not need to wait for the backup to be completed.
  • the output is delivered to improve the throughput of the system; the rapid disaster recovery is provided, and the disaster recovery time for network services and database services is faster than the existing technology.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Quality & Reliability (AREA)
  • Computer Security & Cryptography (AREA)
  • Hardware Redundancy (AREA)
  • Health & Medical Sciences (AREA)
  • Cardiology (AREA)
  • General Health & Medical Sciences (AREA)

Abstract

本发明提供一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法，双机热备份容灾系统包括主服务器与备份服务器，主服务器与备份服务器通过网络连接，主服务器上运行主虚拟机，备份服务器上运行后备虚拟机，后备虚拟机处于主虚拟机的应用层语义上的替代状态，应用层语义上的替代状态是指在应用层语义上后备虚拟机能够代替主虚拟机进行服务，对任何客户请求产生正确的输出。通过可替代性规则比较主虚拟机与后备虚拟机的输出，判断是否需要进行备份，从而有效的降低了备份频率，在保证快速恢复的基础上提升了系统的性能，本发明极大降低了系统开销，提高了系统的吞吐量。

Description

一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法
技术领域
本发明涉及虚拟化环境下的高可靠容灾技术，具体地，涉及一种 虚拟化环境下针对网络服务的双机热备份容灾系统及其方法 。
背景技术
当前，网络化服务是云计算与数据中心的主要服务形式。然而，由于断电、机器硬件故障、灾难或者人为因素（统称故障）的影响，使得这些网络应用停止对外服务、丢失数据的情况时有发生，不仅影响用户的使用，还会带来经济上的损失。因此，如何提高网络服务器的容灾，在故障后快速恢复对外的服务成了许多学者和公司的研究热点。
在现有的研究成果与产品中，有一部分是在虚拟化环境下实现的。
随着计算机技术，特别是网络技术的快速发展和广泛应用，人们对软件的可移植性，特别是软件在网络中移植的需求日益迫切，软件兼容性和可移植性也变得越来越重要。但由于计算机的历史发展过程中，产生了众多不同的、不相兼容的操作系统和指令集架构（Instruction Set Architecture，ISA），从而导致了软件的可移植性被局限在相似的平台上。而在一个大型网络中可能包含基于各种ISA和操作系统的计算机，这就造成了软件可移植性需求和现状之间的矛盾越来越尖锐。而虚拟机（Virtual machine，VM）技术的出现消除了对软件运行平台的这些限制，从而有可能提供更高程度的兼容性和可移植性。虚拟机技术通过在硬件执行平台上加一层软件，从而屏蔽掉平台的差异性，或者说在一个平台上模拟出另一个平台或者多个平台。
目前，基于虚拟机技术的容灾解决方案可以分为检查点（checkpointing）技术和步锁（lockstepping）技术。
Checkpointing 技术利用两台物理设备形成主服务器/后备服务器模式，对同一个应用/虚拟机进行备份，通过虚拟机的迁移技术，定期的将主服务器虚拟机状态备份到后备服务器中，从而实现灾备。后备服务器的虚拟机处于非运行状态，在主服务器故障后能快速恢复到主服务器的前一状态，并继续保留有原先的所有网络连接，这样客户端感觉不到服务端发生了故障且从故障中恢复。但是，为了保证虚拟机间的状态一致性，必须进行周期性的频繁备份（20-40ms一次），导致主服务器的吞吐量会极大降低，CPU开销过大。同时Checkpointing技术将所有服务器发送到客户端的数据包保存在一个缓冲中，只有当一次备份完成后，才能释放这些数据包，这导致网络延迟的增加。
Lockstepping 技术采用双机并行运行的方式保证主服务器后备服务器的状态一致，使得在主服务器故障后客户端能直接与后备服务器连接，快速从故障中恢复。但是Lockstepping技术只能适用于给虚拟机分配单个处理器的情况，对于多处理器虚拟机的性能扩展性很差，多于双处理器的虚拟机性能降低至单处理器是的1/7。另外，对于确定的指令，主后备服务器的虚拟机可以直接并行运行，而对于非确定的指令则需要在主后备服务器虚拟机间做指令级别的同步，这也增加了系统的开销。
发明内容
有鉴于现有技术的上述缺陷，本发明提供一种双机热备份容灾系统。该方案中主虚拟机和后备虚拟机并行运行，根据客户端发送的请求生成各自的输出结果，比较主虚拟机和后备虚拟机的输出结果，如果不一致，需要进行备份，这样既保证了故障后的快速恢复，又有效的降低了系统开销。
本发明提供 一种双机热备份容灾系统，用于虚拟化环境下的网络服务，双机热备份容灾系统包括主服务器与备份服务器，主服务器与备份服务器通过网络连接，其特征在于：主服务器上运行主虚拟机，备份服务器上运行后备虚拟机，后备虚拟机处于主虚拟机的应用层语义上的替代状态，应用层语义上的替代状态是指在应用层语义上后备虚拟机能够代替主虚拟机进行服务，对任何客户请求产生正确的输出。
进一步地，主服务器将客户端请求分别发送给主虚拟机与后备虚拟机，主虚拟机与后备虚拟机并行运行生成各自的响应数据包。
进一步地，双机热备份容灾系统还包括运行在主虚拟机上的主备份管理器，以及运行在后备虚拟机上的后备备份管理器，后备备份管理器用于将后备虚拟机生成的响应数据包发送到主备份管理器，主备份管理器用于比较主虚拟机与后备虚拟机的响应数据包是否一致，如果一致，后备虚拟机处于主虚拟机的可替代状态，主备份管理器将主虚拟机生成的响应数据包发给客户端；如果不一致，后备虚拟机不处于主虚拟机的可替代状态 。
进一步地，如果后备虚拟机不处于主虚拟机的可替代状态，主备份管理器将主虚拟机的当前状 态备份到后备虚拟机。
进一步地 ，备份为 非周期备份。
进一步地 ，备份到后备虚拟机为增量 备份。
系统中采用了增量备份方式，以降低状态备份的开销。不同于现有的检查点技术，本发明中采用双机并行运行，因此在两次状态备份间，后备虚拟机的状态也会发生变化，这导致只备份主虚拟机状态增量是不够的。为了减少备份期间的传送内容，本发明中采用以空间换时间的方法。在主虚拟机与后备虚拟机第一次建立连接时，将主虚拟机的状态完全传输给后备虚拟机，并同时存入后备服务器的一个临时缓存中。每次备份主虚拟机状态时，仅传输在上次备份后发生改变的内容。首先将这些内容更新入后备服务器临时缓存中，然后将临时缓存中内容全备份入后备虚拟机，这就避免了两次备份间后备虚拟机状态变化对增量备份所产生的影响。
进一步地，后备备份管理器检测 主虚 拟机的心跳数据包，如果后备备份管理器没有接收到主虚拟机的心跳数据包，客户端请求数据包直接到达后备虚拟机，后备虚拟机产生响应数据包之后，后备备份管理器将 响应数据包直接发给客户端。
系统中引入心跳包机制，用于后备虚拟机监测主虚拟机是否仍继续存活。如果后备虚拟机没有接收到心跳数据包，就认为主虚拟机已经故障，将采取故障恢复措施以替代主虚拟机继续提供服务。在这种情况下，客户端发送过来的请求数据包将直接到达后备虚拟机，后备虚拟机产生响应数据包之后，不再发送给主虚拟机，而是直接发给客户端。该情况下，客户端收到的数据包来源从主虚拟机改为后备虚拟机，并不会发现服务端进行了一次快速的故障恢复。
进一步地， 在内存备份方面，启用了虚拟机监控器提供的影子页表机制，获取在上次状态备份之后被修改的页面。其基本原理是将所有虚拟机的页面都改成写保护，这样，一旦对某个页面进行写操作，就会触发一个异常并进入异常处理程序。
本发明还提供 一种双机热备份容灾方法，其特征在于，包括以下步骤：
（1） 主服务器通过流量控制，将客户端发出的请求分别发送到主虚拟机和后备虚拟机；
（2）主虚拟机与后备虚拟机根据客户请求并行运行，生成各自的响应数据包；
（ 3 ） 后备备份管理器将后备虚拟机生成的响应数据包发送到主备份管理器；
（4） 主备份管理器用于比较主虚拟机与后备虚拟机的响应数据包是否一致，如果一致，后备虚拟机处于主虚拟机的应用层语义上的替代状态， 将主虚拟机的响应数据包发给客户端 ；如果不一致，后备虚拟机不处于主虚拟机的应用层语义上的替代状态 ， 主备份管理器将主虚拟机的当前状 态备份到后备虚拟机。
与现有技术相比，本发明提供的双机热备份容灾系统及其方法， 具有以下有益的技术效果：
（1）系统实现中解决了主后备服务器双机并行情况下，存储访问的一致性，网络协议的一致性，多核状态CPU指令一致性等技术问题。
（2） 基于可替代性规则，本方案中主服务器状态的备份为非周期性，备份间隔大于1秒，频率相对于现有技术降低两个数量级以上，这极大降低了系统开销，基本消除虚拟机状态备份对主服务器的性能干扰。
（3）相比于现有解决方案，本发明中主服务器不需等待备份完成即可交付输出结果，提高了系统的吞吐量。
（4）本发明方案能够提供快速灾备恢复，对于网络服务、数据库服务的灾备恢复时间快于现有技术。
附图说明
图1是现有检查点技术的流程示意图；
图2是现有步锁技术的流程示意图；
图3是本发明的一个实施例的双机热备份容灾系统的流程示意图；
图4是本发明的一个实施例的双机热备份容灾系统的增量备份的过程示意图。
具体实施方式
以下将结合附图对本发明的构思、具体结构及产生的技术效果作进一步说明，以充分地了解本发明的目的、特征和效果。
图1是现有检查点技术的流程示意图。主虚拟机处理客户端请求并产生响应，后备虚拟机则处于非运行状态。在主服务器中有定时模块产生周期性事件，备份管理器在收到事件后获取主虚拟机状态，将在上一次备份后改变的状态备份到后备虚拟机上。
图2是现有步锁技术的流程示意图。主虚拟机与后备虚拟机并行执行客户端发来的请求，主虚拟机将响应发回到客户端。由于非确定的指令（例如内存访问、时钟中断等）则需要在虚拟机间做指令级别的同步，以避免两边的状态发生差异。
本发明提供一种双机热备份容灾系统，用于虚拟化环境下的网络服务，双机热备份容灾系统包括主服务器与备份服务器，主服务器与备份服务器通过网络连接，其特征在于：主服务器上运行主虚拟机，备份服务器上运行后备虚拟机，后备虚拟机处于主虚拟机的应用层语义上的替代状态，应用层语义上的替代状态是指在应用层语义上后备虚拟机能够代替主虚拟机进行服务，对任何客户请求产生正确的输出。
客户端发送过来的请求数据包首先抵达外围交换机，交换机通过目的MAC地址来决定转发端口。当主虚拟机提供服务时，交换机学习到的虚拟机MAC地址对应端口为主服务器网卡端口，因此请求数据包被发往主服务器。
主服务器将客户端请求分别发送给主虚拟机与后备虚拟机，主虚拟机与后备虚拟机并行运行生成各自的响应数据包。
双机热备份容灾系统还包括运行在主虚拟机上的主备份管理器，以及运行在后备虚拟机上的后备备份管理器，后备备份管理器用于将后备虚拟机生成的响应数据包发送到主备份管理器，主备份管理器用于比较主虚拟机与后备虚拟机的响应数据包是否一致，如果一致，后备虚拟机处于主虚拟机的可替代状态，主备份管理器将主虚拟机生成的响应数据包发给客户端；如果不一致，后备虚拟机不处于主虚拟机的可替代状态 。
如果后备虚拟机不处于主虚拟机的可替代状态，主备份管理器将主虚拟机的当前状 态备份到后备虚拟机。
备份为非周期备份。
备份到后备虚拟机为增量备份。
系统中采用了增量备份方式，以降低状态备份的开销。不同于现有的检查点技术，本发明中采用双机并行运行，因此在两次状态备份间，后备虚拟机的状态也会发生变化，这导致只备份主虚拟机状态增量是不够的。为了减少备份期间的传送内容，本发明中采用以空间换时间的方法。在主虚拟机与后备虚拟机第一次建立连接时，将主虚拟机的状态完全传输给后备虚拟机，并同时存入后备服务器的一个临时缓存中。每次备份主虚拟机状态时，仅传输在上次备份后发生改变的内容。首先将这些内容更新入后备服务器临时缓存中，然后将临时缓存中内容全备份入后备虚拟机，这就避免了两次备份间后备虚拟机状态变化对增量备份所产生的影响。
后备备份管理器检测 主虚 拟机的心跳数据包，如果后备备份管理器没有接收到主虚拟机的心跳数据包，客户端请求数据包直接到达后备虚拟机，后备虚拟机产生响应数据包之后，后备备份管理器将 响应数据包直接发给客户端。
系统中引入心跳包机制，用于后备虚拟机监测主虚拟机是否仍继续存活。如果后备虚拟机没有接收到心跳数据包，就认为主虚拟机已经故障，将采取故障恢复措施以替代主虚拟机继续提供服务。 后备服务器将会向交换机发送一个 ARP 数据包，该数据包的源 MAC 地址就是后备虚拟机的 MAC 地址。这使得交换机学习到一个新的 MAC 地址到端口的映射条目。之后客户端发送过来的目的 MAC 地址是虚拟机的数据包，将直接发送给后备服务器的网卡。 后备虚拟机产生响应数据包之后，不再发送给主虚拟机，而是直接发给客户端。该情况下，客户端收到的数据包来源从主虚拟机改为后备虚拟机，并不会发现服务端进行了一次快速的故障恢复。
在内存备份方面，启用了虚拟机监控器提供的影子页表机制。获取在上次状态备份之后哪些页面被修改。其基本原理是将所有虚拟机的页面都改成写保护，这样，一旦对某个页面进行写操作，就会触发一个异常并进入异常处理程序。借助于'影子页表'机制，能够很容易获取在上次状态备份之后，哪些页面被修改。
图3是本实施例的双机热备份容灾系统的流程示意图，具体过程如下：
步骤1、主服务器将客户端发出的请求包分发到主虚拟机与后备虚拟机，过程如下：首先，客户端发送过来的请求数据包通过外围交换机被交换机发送到主服务器。主服务器接收数据包后送到软件网桥，在软件网桥中通过配置Linux自带的工具TC（Traffic Control）来实现对网络数据包的截获和分发，将数据包发送到主虚拟机与后备虚拟机。
TC 的配置方法如下：
#tc qdisc add dev vif1.0 root handle 1: prio
#tc filter add dev vif1.0 parent 1: protocol ip prio 10 u32 match u32 0 0 flowid 1:2 action mirred egress mirror dev eth0
#tc filter add dev vif1.0 parent 1: protocol arp prio 11 u32 match u32 0 0 flowid 1:2 action mirred egress mirror dev eth0
步骤2、主虚拟机与后备虚拟机根据应用层语义并行执行，生成各自的输出，后备虚拟机将输出发送到主服务器。通过配置TC实现后备虚拟机输出的拦截与转发，具体方法如下：
#tc qdisc add dev vif1.0 ingress
#tc filter add dev vif1.0 parent ffff: protocol ip prio 10 u32 match u32 0 0 flowid 1:2 action mirred egress redirect dev eth0
步骤3、主服务器的管理器比较主虚拟机与后备虚拟机生成各自的输出是否满足可替代性规则。具体的，在管理器中实现两个队列形式的虚拟接口，将主虚拟机与后备虚拟机的输出分别重定向入一个接口。管理器通过对两队列中的包进行逐一比较，判断后备虚拟机是否仍是主虚拟机的可替代状态。通过配置TC实现输出的重定向，具体方法如下：
a ）主虚拟机输出包重定向：
#tc qdisc add dev vif1.0 ingress
#tc filter add dev vif1.0 parent ffff: protocol ip prio 10 u32 match u32 0 0 flowid 1:2 action mirred egress redirect dev ifb0
b ）后备虚拟机输出包重定向：
#tc qdisc add dev eth0 ingress
#tc filter add dev eth0 parent ffff: protocol ip prio 10 u32 match u32 0 0 flowid 1:2 action mirred egress redirect dev ifb1
步骤4、将主服务器的输出作为响应包发回给客户端。
步骤5、若判断后备虚拟机不是主虚拟机的可替代状态，则将主虚拟机的当前状态备份到后备虚拟机。在主服务器上和后备服务器的管理器中上都有一个备份守护进程，负责对状态的发送、接收和对虚拟机状态的更新。
图4是本实施例的双机热备份容灾系统的增量备份的过程示意图。
步骤1、主服务器上的备份管理器获取主虚拟机在上次备份后状态变更部分。
步骤2、备份管理器将变更部分发送到后备虚拟机。
步骤3、后备虚拟机将变更部分更新临时缓存。
步骤4、将临时缓存内容全部备份入后备虚拟机。
在磁盘文件备份方面，通过修改磁盘设备的后端驱动，截获主虚拟机与后备虚拟机对磁盘的写入操作。在两次备份间主虚拟机与后备虚拟机的磁盘写入数据都被暂时保存在各自的临时缓存中。当备份时用主虚拟机临时缓存中内容替换备份虚拟机临时缓存的内容，然后分别写入磁盘。
在设备备份方面，由于设备状态涉及虚拟机监控器的前后端模型，获取较困难，因此选用丢弃主虚拟机与后备虚拟机设备驱动之前的状态。当备份完成后重新建立连接来使得设备状态保持一致。
本发明提供 的双机热备份容灾系统及其方法， 解决了主后备服务器双机并行情况下，存储访问的一致性，网络协议的一致性，多核状态CPU指令一致性等技术问题；基于可替代性规则，本方案中主服务器状态的备份为非周期性，备份间隔大于1秒，频率相对于现有技术降低两个数量级以上，这极大降低了系统开销，基本消除虚拟机状态备份对主服务器的性能干扰；主服务器不需等待备份完成即可交付输出结果，提高了系统的吞吐量；提供快速灾备恢复，对于网络服务、数据库服务的灾备恢复时间快于现有技术。
以上详细描述了本发明的较佳具体实施例。应当理解，本领域的普通技术人员无需创造性劳动就可以根据本发明的构思做出诸多修改和变化。因此，凡本技术领域中技术人员依本发明的构思在现有技术的基础上通过逻辑分析、推理或者有限的实验可以得到的技术方案，皆应在由权利要求书所确定的保护范围内。

Claims (9)

  1. 一种双机热备份容灾系统，用于虚拟化环境下的网络服务，所述双机热备份容灾系统包括主服务器与备份服务器，所述主服务器与所述备份服务器通过网络连接，其特征在于，所述主服务器上运行主虚拟机，所述备份服务器上运行后备虚拟机，所述后备虚拟机处于所述主虚拟机的应用层语义上的替代状态，所述应用层语义上的替代状态是指在应用层语义上后备虚拟机能够代替主虚拟机进行服务，对任何客户请求产生正确的输出。
  2. 如权利要求1所述的双机热备份容灾系统，其特征在于，所述主服务器将客户端请求分别发送给所述主虚拟机与所述后备虚拟机，所述主虚拟机与所述后备虚拟机并行运行生成各自的响应数据包。
  3. 如权利要求2所述的双机热备份容灾系统，其特征在于，所述双机热备份容灾系统还包括运行在所述主虚拟机上的主备份管理器，以及运行在所述后备虚拟机上的后备备份管理器，所述后备备份管理器用于将所述后备虚拟机生成的所述响应数据包发送到所述主备份管理器，所述主备份管理器用于比较所述主虚拟机与所述后备虚拟机的响应数据包是否一致，如果一致，所述后备虚拟机处于所述主虚拟机的可替代状态，所述主备份管理器将所述主虚拟机生成的所述响应数据包发给所述客户端；如果不一致，所述后备虚拟机不处于所述主虚拟机的可替代状态。
  4. 如权利要求3所述的双机热备份容灾系统，其特征在于，如果所述后备虚拟机不处于所述主虚拟机的可替代状态，所述主备份管理器将所述主虚拟机的当前状态备份到所述后备虚拟机。
  5. 如权利要求4所述的双机热备份容灾系统，其特征在于，所述备份为非周期备份。
  6. 如权利要求4所述的双机热备份容灾系统，其特征在于，所述备份到所述后备虚拟机为增量备份。
  7. 如权利要求3所述的双机热备份容灾系统，其特征在于，所述后备备份管理器检测所述主虚拟机的心跳数据包，如果所述后备备份管理器没有接收到所述主虚拟机的心跳数据包，所述后备虚拟机产生响应数据包之后，所述后备备份管理器将所述响应数据包直接发给所述客户端。
  8. 如权利要求1所述的双机热备份容灾系统，其特征在于，在内存备份方面，启用了虚拟机监控器提供的影子页表机制，获取在上次状态备份之后被修改的页面。
  9. 一种如权利要求1-8所述的双机热备份容灾系统的双机热备份容灾方法，其特征在于，包括以下步骤：
    （1）主服务器通过流量控制，将客户端发出的请求分别发送到主虚拟机和后备虚拟机；
    （2）所述主虚拟机与所述后备虚拟机根据所述客户端发出的请求并行运行，生成各自的响应数据包；
    （3）后备备份管理器将所述后备虚拟机生成的所述响应数据包发送到主备份管理器；
    （4）所述主备份管理器用于比较所述主虚拟机与所述后备虚拟机的响应数据包是否一致，如果一致，所述后备虚拟机处于所述主虚拟机的应用层语义上的替代状态，将所述主虚拟机的响应数据包发给所述客户端；如果不一致，所述后备虚拟机不处于所述主虚拟机的应用层语义上的替代状态，所述主备份管理器将所述主虚拟机的当前状态备份到所述后备虚拟机。
PCT/CN2014/083113 2014-01-22 2014-07-28 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法 Ceased WO2015109804A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US14/412,125 US20160323427A1 (en) 2014-01-22 2014-07-28 A dual-machine hot standby disaster tolerance system and method for network services in virtualilzed environment

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410029760.5 2014-01-22
CN201410029760.5A CN103761166A (zh) 2014-01-22 2014-01-22 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法

Publications (1)

Publication Number Publication Date
WO2015109804A1 true WO2015109804A1 (zh) 2015-07-30

Family

ID=50528408

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/083113 Ceased WO2015109804A1 (zh) 2014-01-22 2014-07-28 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法

Country Status (3)

Country Link
US (1) US20160323427A1 (zh)
CN (1) CN103761166A (zh)
WO (1) WO2015109804A1 (zh)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10209981B2 (en) 2016-11-21 2019-02-19 Nio Usa, Inc. Structure for updating software in remote device
US10360020B2 (en) * 2017-04-11 2019-07-23 Nio Usa, Inc. Virtual machine (VM) approach to embedded system hot update
US10871952B2 (en) 2017-12-20 2020-12-22 Nio Usa, Inc. Method and system for providing secure over-the-air vehicle updates
US10887382B2 (en) 2018-12-18 2021-01-05 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
CN112202594A (zh) * 2020-09-07 2021-01-08 核电运行研究(上海)有限公司 一种核电站服务器故障应急处理系统及方法
US10958720B2 (en) 2018-12-18 2021-03-23 Storage Engine, Inc. Methods, apparatuses and systems for cloud based disaster recovery
US10983886B2 (en) 2018-12-18 2021-04-20 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11176002B2 (en) 2018-12-18 2021-11-16 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11178221B2 (en) 2018-12-18 2021-11-16 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11252019B2 (en) 2018-12-18 2022-02-15 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11489730B2 (en) 2018-12-18 2022-11-01 Storage Engine, Inc. Methods, apparatuses and systems for configuring a network environment for a server

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103761166A (zh) * 2014-01-22 2014-04-30 上海交通大学 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法
CN105577711B (zh) * 2014-10-08 2019-05-03 华为技术有限公司 消息处理方法、装置及消息处理系统
CN104618155B (zh) * 2015-01-23 2018-06-05 华为技术有限公司 一种虚拟机容错的方法、装置及系统
CN104899071A (zh) * 2015-04-29 2015-09-09 深圳市深信服电子科技有限公司 集群中虚拟机的恢复方法和系统
CN105119754A (zh) * 2015-09-08 2015-12-02 烽火通信科技股份有限公司 虚拟化主备倒换保持tcp连接的系统及方法
CN105656670B (zh) * 2015-12-31 2019-08-23 北京航管软件技术有限公司 多控制卡线路切换装置及其控制方法
CN108259629B (zh) * 2016-12-28 2021-07-23 阿里巴巴集团控股有限公司 虚拟互联网协议地址的切换方法及装置
CN107315660A (zh) * 2017-06-29 2017-11-03 郑州云海信息技术有限公司 一种虚拟化系统的双机热备份方法、装置及系统
CN107656845A (zh) * 2017-09-18 2018-02-02 国云科技股份有限公司 一种虚拟机高可用方法
CN109561025B (zh) * 2017-09-27 2022-04-05 华为技术有限公司 一种信息处理方法以及相关设备
CN109240799B (zh) * 2018-09-06 2022-04-15 福建星瑞格软件有限公司 大数据平台集群容灾方法、系统及计算机可读存储介质
CN109460314B (zh) * 2018-11-13 2022-02-11 天津津航计算技术研究所 一种嵌入式系统的双机热备装置
CN109271274B (zh) * 2018-11-13 2022-02-11 天津津航计算技术研究所 一种嵌入式系统的双机热备方法
CN110062057A (zh) * 2018-12-18 2019-07-26 华为技术有限公司 为热备系统处理报文的代理网关及通信方法
CN110515763A (zh) * 2019-07-26 2019-11-29 浪潮电子信息产业股份有限公司 一种基于OpenStack的虚拟机双机热备的方法和系统
CN110727733B (zh) * 2019-09-25 2023-07-14 许昌许继软件技术有限公司 一种主备服务器系统及数据同步方法
CN111371625A (zh) * 2020-03-18 2020-07-03 北京佳讯飞鸿电气股份有限公司 一种双机热备的实现方法
CN114095964B (zh) * 2021-11-19 2023-05-26 中国联合网络通信集团有限公司 一种故障恢复方法、装置及计算机可读存储介质
CN117041320A (zh) * 2023-08-18 2023-11-10 郑州云海信息技术有限公司 一种数据传输方法、装置、设备及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100318991A1 (en) * 2009-06-15 2010-12-16 Vmware, Inc. Virtual Machine Fault Tolerance
CN103412800A (zh) * 2013-08-05 2013-11-27 华为技术有限公司 一种虚拟机热备份方法和设备
CN103501290A (zh) * 2013-09-18 2014-01-08 万达信息股份有限公司 一种基于动态备份虚拟机的高可靠服务系统构建方法
CN103761166A (zh) * 2014-01-22 2014-04-30 上海交通大学 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6397242B1 (en) * 1998-05-15 2002-05-28 Vmware, Inc. Virtualization system including a virtual machine monitor for a computer with a segmented architecture
US20030177149A1 (en) * 2002-03-18 2003-09-18 Coombs David Lawrence System and method for data backup

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100318991A1 (en) * 2009-06-15 2010-12-16 Vmware, Inc. Virtual Machine Fault Tolerance
CN103412800A (zh) * 2013-08-05 2013-11-27 华为技术有限公司 一种虚拟机热备份方法和设备
CN103501290A (zh) * 2013-09-18 2014-01-08 万达信息股份有限公司 一种基于动态备份虚拟机的高可靠服务系统构建方法
CN103761166A (zh) * 2014-01-22 2014-04-30 上海交通大学 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10209981B2 (en) 2016-11-21 2019-02-19 Nio Usa, Inc. Structure for updating software in remote device
US10360020B2 (en) * 2017-04-11 2019-07-23 Nio Usa, Inc. Virtual machine (VM) approach to embedded system hot update
US10871952B2 (en) 2017-12-20 2020-12-22 Nio Usa, Inc. Method and system for providing secure over-the-air vehicle updates
US10887382B2 (en) 2018-12-18 2021-01-05 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US10958720B2 (en) 2018-12-18 2021-03-23 Storage Engine, Inc. Methods, apparatuses and systems for cloud based disaster recovery
US10983886B2 (en) 2018-12-18 2021-04-20 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11176002B2 (en) 2018-12-18 2021-11-16 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11178221B2 (en) 2018-12-18 2021-11-16 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11252019B2 (en) 2018-12-18 2022-02-15 Storage Engine, Inc. Methods, apparatuses and systems for cloud-based disaster recovery
US11489730B2 (en) 2018-12-18 2022-11-01 Storage Engine, Inc. Methods, apparatuses and systems for configuring a network environment for a server
CN112202594A (zh) * 2020-09-07 2021-01-08 核电运行研究(上海)有限公司 一种核电站服务器故障应急处理系统及方法

Also Published As

Publication number Publication date
US20160323427A1 (en) 2016-11-03
CN103761166A (zh) 2014-04-30

Similar Documents

Publication Publication Date Title
WO2015109804A1 (zh) 一种虚拟化环境下针对网络服务的双机热备份容灾系统及其方法
US8307363B2 (en) Virtual machine system, restarting method of virtual machine and system
US8694828B2 (en) Using virtual machine cloning to create a backup virtual machine in a fault tolerant system
US9971660B2 (en) Virtual machine network loss detection and recovery for high availability
US9690662B2 (en) Techniques for virtual machine management
US9641627B2 (en) Techniques for remapping sessions for a multi-threaded application
US8069218B1 (en) System, method and computer program product for process migration with planned minimized down-time
JP3844621B2 (ja) アプリケーション実現方法及びアプリケーション実現装置
Reiser et al. Hypervisor-based efficient proactive recovery
JP5817308B2 (ja) サーバ、サーバシステムおよびサーバの冗長切り替え方法
CN102662751A (zh) 一种提高基于热迁移虚拟机系统可用性的方法
WO2018058942A1 (zh) 一种数据处理方法以及备份服务器
WO2015180434A1 (zh) 一种数据库集群管理数据的方法、节点及系统
US20100180094A1 (en) Storage system, backup storage apparatus, and backup control method
WO2018120680A1 (zh) 虚拟磁盘备份系统、方法、装置、服务主机和存储介质
WO2021051492A1 (zh) 数据库服务节点切换方法、装置、设备及计算机存储介质
WO2017157125A1 (zh) 在云计算环境中删除云主机的方法、装置、服务器及存储介质
JP5773166B2 (ja) 計算機の制御方法、計算機及び計算機システム
US10241874B2 (en) Checkpoint method for a highly available computer system
Ong et al. VCCP: A transparent, coordinated checkpointing system for virtualization-based cluster computing
TWI763331B (zh) 虛擬機器的備用方法與備用系統
Yaothanee et al. An in-memory checkpoint-restart mechanism for a cluster of virtual machines
Wang et al. Design of a live networking device update mechanism for cloud computing systems
CN107423113B (zh) 一种管理虚拟设备的方法、带外管理设备及备用虚拟设备
Wang et al. A remote backup approach for virtual machine images

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 14412125

Country of ref document: US

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

Ref document number: 14879845

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 25/11/2016)

122 Ep: pct application non-entry in european phase

Ref document number: 14879845

Country of ref document: EP

Kind code of ref document: A1