WO2014146388A1 - 一种分组设备运行状态同步方法 - Google Patents

一种分组设备运行状态同步方法 Download PDF

Info

Publication number
WO2014146388A1
WO2014146388A1 PCT/CN2013/079953 CN2013079953W WO2014146388A1 WO 2014146388 A1 WO2014146388 A1 WO 2014146388A1 CN 2013079953 W CN2013079953 W CN 2013079953W WO 2014146388 A1 WO2014146388 A1 WO 2014146388A1
Authority
WO
WIPO (PCT)
Prior art keywords
running status
block
entry
network management
status block
Prior art date
Application number
PCT/CN2013/079953
Other languages
English (en)
French (fr)
Inventor
马红斌
张斯斯
Original Assignee
烽火通信科技股份有限公司
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 烽火通信科技股份有限公司 filed Critical 烽火通信科技股份有限公司
Publication of WO2014146388A1 publication Critical patent/WO2014146388A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks

Definitions

  • the invention relates to the running state of the network management and the grouping device in the field of packet communication, and is specifically a method for synchronizing the running state of the grouping device. Background technique
  • services such as Layer 2 VPN (Virtual Private Network), Layer 3 paste, and LSP (Label Switch Path) are deployed by the NMS to the device and are controlled on the device side.
  • the plane protocol is dynamically distributed, the corresponding service can be established.
  • the status of the service changes, and the service state is established after the service is dynamically established or the service state changes.
  • This state is not a single-state that contains only a small amount of data. It contains a large amount of dynamic service information. It needs to be reported to the network management system in time to maintain the synchronization of the network management and the running status of the device (Synchronous). After the NMS or device is restarted, the NMS needs to obtain the running status of the device in time to maintain synchronization.
  • the same status of the running state uses the full amount of the same method, all the contents of the same state each time.
  • the number of backups the number of NEs
  • the number of peers in the running state is very large, which is both inefficient and wastes system resources. In many cases, this function cannot be used.
  • large packet networks the number of network elements is several thousand
  • the efficiency of the same state of operation is crucial, and this problem needs to be solved urgently.
  • the object of the present invention is to provide a method for synchronizing the running state of a grouping device, combining the increments with the full amount of peers, and adopting a data organization method that separates the summary from the specific data.
  • the running state synchronization is well completed, and the accuracy of the operating state can be ensured while improving the synchronization efficiency of the running state.
  • the technical solution adopted by the present invention is:
  • a grouping device running state synchronization method is characterized in that: when a packet device carries different types of services, an operating state of a service of a certain service corresponds to a running status block entry RS J , a content of a running status block entry RS J Packetcast: Entry Summary RS-S and Entry Data,
  • the running status block entry RS-I of all the services of the certain service and the summary of the block RS-S constitute the running status block RS of the service.
  • the synchronous summary refers to; Or the summary RS-S of the synchronization block; the synchronization content refers to: the number of synchronization entries, or the number of synchronization blocks, the number of blocks refers to the set of running status block entries RS-I of all services of a certain service .
  • the device side when the device side incrementally reports, only the running status entry that changes on the device side is reported, and the entry summary RSJ-S of the changed running status block entry RS-I is reported to the network management incrementally ; After receiving the reported item summary RS_I-S, the item data corresponding to the running status block entry RS ⁇ in the report is retrieved from the device side.
  • the summary RSS and the entry summary RS-I-S of the synchronization block are compared, and the difference between the network management side and the device side is compared, and only the different parts of the network management side and the device side are synchronized. .
  • Step (1) the network management acquires the summary RS-S of the running state block RS one by one from the device side, and the operation of the network management cache The summary KS of the block of the status block RS. For comparison, if it is consistent, the summary RS_S of the block of the next operational status block RS is obtained, and the above ratio «, until the sum of the blocks of all the operating status blocks RS is repeated « Finished; if they are inconsistent, perform step (2);
  • Step (2) Extracting the summary RS ⁇ LS of all the entries of the running status block RS from the device side, and then comparing with the summary RS-I-S of the running status block RS saved on the network management one by one, if they are consistent, Get the next one and repeat the above comparison until all The summary RSJLS of the partial entries is compared; if they are inconsistent, step (3) is performed ; and step (3) is taken from the device side to retrieve the contents of the RS J corresponding to the running status entry.
  • the running state of the grouping device captured by the invention is the same as the method of combining the incremental synchronization and the full amount of the same, and the data organization method for separating the abstract and the specific data is used, and the running state synchronization can be well completed, and the running is improved.
  • the state synchronization efficiency also ensures the accuracy of the operating state.
  • the invention has the following appendix:
  • FIG. 1 is an organization structure of a packet device running status block RS according to the present invention
  • FIG. 2 is a packet device running status block reporting buffer RS R (Run) provided by the present invention.
  • FIG. 3 is a schematic diagram of the W tube and equipment operating state organization provided by the present invention.
  • ⁇ 4 provides RSJJJNCILREPORT message knot for the present invention
  • Figure 5 is a structure of an RS-IJffiCLRUN STATUS message provided by the present invention.
  • ⁇ 10 is the NE-STARTUPJffiPORT message structure provided by the present invention.
  • FIG. 11 is a schematic diagram of an NE-STARTUP_REPORT_C0NF0 ⁇ message structure provided by the present invention
  • FIG. 12 is a diagram showing an interaction process of a device incremental state reporting device in the same manner as the device provided in the present invention.
  • FIG. 13 is a diagram showing an exchange process of a network controller actively executing a full amount of peers in a device running state synchronization scheme provided by the present invention
  • the circle 14 is a schematic diagram of the full-scale synchronization process of the network management after the device-side restart of the device running state synchronization scheme provided by the present invention. Specific way
  • the method for synchronizing the running state of the grouping device provides an effective scheme for the running state of the grouping device, which comprises the following four parts as a whole: the organization method of the data (the separation of the data and the data), the first The combination of the same content strategy, the backup side incremental reporting and the network management initiative to perform the full amount of synchronization combined strategy.
  • the method for synchronizing the running status of the grouping device of the present invention includes the following steps: Step 1: Separating the summary and the data by the organization of the running state content of the service, the two categories of the summary and the data Performing the separation means: dividing the running status block RS for describing the running status of the service into the summary RS_S of the running status block (referred to as the summary of the block or the data of the running status block (referred to as the number of blocks), the block
  • the data is a set of several operating status block entries RS I (referred to as entries RS I or RS I or entries). Therefore, the summary and the number of two categories are separated, and each running status block entry RS-I is divided. For the summary of the article RSJ-S (referred to as RS_LS) and the number of entries; delete 1;
  • a packet device carries thousands of different types of services, such as LSPs, Layer 2 leaks, and Layer 3 VPNs.
  • One of the service operations of a certain service corresponds to a running status block entry RSJ (Run Status Item).
  • the content package of a running status block entry RS J The entry summary RS-I-S (Run Status Item Sumiiary, RS-I-S) and the entry data, the entry summary RS_LS and the entry data are determined according to the type of service, obviously different The business has different entry summaries and entry data.
  • the set of running status block entries RS_I of all the services of the certain service is used as the block data, and the Run Status BLOCK Summary (RS_S) of the block constitutes the running status block RS of the service.
  • R_S Run Status BLOCK Summary
  • RS-B Run Status BLOCK, RS-B. That is: the content of a running status block RS includes the summary of the block RS_S and the data of the block, and the data of the block refers to the set of all running status block bars RS i of the block.
  • Step 2 Adopt a strategy of synchronizing the content and synchronizing the content first, that is, synchronizing the summary and then Step content, whether the network management actively performs full synchronization or device-side incremental reporting (incremental synchronization), whether it is a synchronous running status bar or a synchronous running status block, is a policy that follows the content of the first summary.
  • the order of the peers of the block (referring to the running status block RS) is the content of the block's summary RS S. block (ie the block's data).
  • Synchronization order of entries Summary of entries
  • Step 3 the device side increments «
  • the device-side incremental report only reports the running status entries that have changed on the device side. That is, when a part of the running status block entry RS_I of a running status block on the device side changes, the item summary RS_I_S of the changed running status block entry RS_I is reported to the network management incrementally; the received item is received. After the digest, the item data corresponding to the running status block entry RSJ in the report is retrieved from the device side.
  • Step 4 The network administrator actively executes the full amount of peers.
  • the NMS initiates full-scale synchronization to the device side.
  • the full-scale synchronization here is not the content of all entries of all running status blocks, but the summary RS-S of the same block.
  • the item summary RS-I-S compares the difference between the network management and the device side, and then only synchronizes different parts of the network management side and the equipment side.
  • the specific synchronization process is as follows:
  • the network management device obtains a summary RS-S of the block of the running status block RS from the device, and compares it with the summary RS-S of the block of the running status block KS of the network management cache. If they are consistent, the following is obtained. a summary RS-S of the block of the running status block RS, repeating the above comparison until all the summary RS_S of the block of the running status block RS is compared; if not, executing step (2);
  • Step (2) extracting, from the device side, a summary RS-I-S of all the running status block RSs, and then comparing one by one with the summary RSJ-S of the running status block RS stored on the network management tube, if consistent, obtaining The next one is repeated and the above comparison is repeated until the summary RS-I-S of all the entries are compared; if not, the step (3) is performed;
  • Step (3) fetching the content of the corresponding running state entry RS-1 from the device side (the content refers to the entry data). After the strip data of a running status bar RS I is obtained, the next running status item RS_I is obtained, and the digest of the item described in the step (2) is continued RS-I-S For comparison, after processing one running status block RS, the next running status block RS is obtained, and the comparison of the summary RS-S of the block described in step (1) is continued.
  • the following is further described by way of examples.
  • FIG. 1 is an organization structure of a type operation status block HS of a packet device.
  • the content of an operation status block RS includes a summary KS_S of a block and data of a block, and the data of the block refers to a set of all running status block entries RSJ of the operation status block.
  • the summary of the block RS S packet insertion type BJ>pe, feature bit BJlag, total number of entries I-Count block sequence number B-Seq, block change time ⁇ B_ Sta*p, block checksum B-CRC, block
  • the B-Seq and BJtamp in the summary RS-S are the inverse of the entire running state block, that is, any change of any one of the running state blocks RS, Seq is increased by one, and i B-Staiip also changes.
  • each running status block entry RSJ includes the item digest RS_I_S and the data of the item, and the item digest includes: an item identifier 1 ⁇ 0), an item sequence number; LS e q, an item flag bit I lag entry checksum LCRC, whose ⁇ Seq reflects the change of the entry, the data of the entry changes once, and I Seq is incremented by one.
  • the entry flag I-flag indicates whether it is added, deleted or updated, and also indicates whether the report requires confirmation.
  • Figure 2 is a type of running status block report buffer of the device ( Run Status
  • Block Report the organization structure of the RSJO, each running status block report buffer RS_R, including the RS S pointer (pointing to the RS S address in the running status block RS),
  • B—SeiCoirf irmed is the confirmation value of the latest block Seq (B-Seq) reported by the network management system. After the network management is restarted, the incremental peers can be continued.
  • I—Report—Co ⁇ t is the number of reported entries.
  • RS-I-S in RS-R is a copy of the corresponding entry summary corresponding to the running status block RS. The device side generates an RS-R for each NMS for each running status block.
  • the RS R includes a summary of the RS-I to be reported.
  • Figure 3 shows the data organization of the network management system and the device in units of running status blocks.
  • a running status block RS stores the contents of a type of running status block (including the summary of the block and the number of blocks)
  • the running status block reports the buffer RS-R, and the buffer stores the RS part information (partially means only the summary RS ⁇ LS of the part items that need to be reported), and the cached entry needs to be reported.
  • a summary of I including entries for which the reported value has not been reported and entries for which the reported letter has not yet received an acknowledgment.
  • the RS R cache device of the network management ⁇ RSJJJNCRJffiPORT message reports the summary of the changed entry, and does not receive the content of the entry; if the content is not received, the entry is never deleted, except ⁇ receives the entry Latest content. If the item is identified
  • the cache of this item is deleted. If different, after comparing I-Seq, replace the I-Seq of the corresponding entry in the buffer with a large I-Seq.
  • the device reports all entries in the RS-R in a timer. After receiving the confirmation from the network administrator, the corresponding entry RSJ-S is deleted from RS-R (not all deleted, preventing new entries between reporting and confirmation). If I-ID-like, delete this entry without judging whether the new I-Seq is consistent.
  • the device has a global start Jiolding flag that controls the processing of the device during the restart.
  • the RS-S in the RS-R is copied to the RS-S field of the message, and each RS in the RS R table is traversed to generate an RSJ-S-INCILREP0RT 3 ⁇ 4 message to the NMS.
  • the network management receives the RS-I-S-INCR-REPORT message, and modifies its own RS R digest, in which the B_Seq in the message is copied to the SeqJ ⁇ nf irmed field of RS R, and the RS-R is added or modified.
  • RS-I-S The NMS maintains the RS R timer timeout and traverses the RS-I-S generation in RS R.
  • the RSJ-REQJDNFI message in which the B JeqJ3 ⁇ 4nf irmed in the RS copies the message B_Seq field, and sends an RS-I-REQ-CONFIRM message to the device side.
  • the Seq field in the message packet modifies the B Jeq-Conf irmed in the RS, traverses each RS JJ entry in the message, and takes the ID of the entry as a key.
  • RS-R table ⁇ finds RS-I-S, if found, deletes the RS-entry from the RS-R table (indicating receipt of the report confirmation); finds RS I from the RS table, if found, in RS-LIES A complete RS-I message is generated for the entry, and if not found, the message is deleted.
  • the RS-I-RES only the RS-I-S digest is generated for the entry, and the LFlag in the digest indicates the special ffi position of the deletion; the RS-I-RES message is sent to the network management.
  • the network management When the RS-I-RES message, it traverses each RS-I entry in the message, and looks up the RS IS from the network management RS R table with the ID of the entry as the key, if found, and receives the Seq If it is larger than the RS R table Seq, delete the RS-I-S entry (indicating that the content is received, no need to retrieve the content of the entry); look up the RS-I from the RS table, if the Seq is found and compared, if the message is in the message If it is not large, modify the RS-I information; if it is not found, add a new RS-I; and modify the RS summary.
  • the NMS sends an RS-SJffiQ message of the running status block to the device side.
  • the NMS After receiving the RS.S. Ri:S message on the device side, the NMS compares the B-Seq and HS t amp of the NMS RS and the message. If they are identical, the full synchronization process is exited; otherwise, the B in the RS is compared. -Seq and Se(LConfirmed) of the message, if they are the same, exit the full synchronization process; otherwise, add RS ⁇ existing RS-I-S to RS-R, and send ALL-RS-I-S-REQ to the device side Text.
  • the device side After receiving the ALL-RS-IJ-REQ message, the device side copies the summary of all RSJs of the RS to RS-R ⁇ . 5) The device side reporting timer expires, traversing the RSJLS RSJJ JNCILREP0RT message in RS R, and then sending the message to the NMS.
  • Subsequent processing refers to the subsequent processing flow of the incremental reporting of the running status block on the device side.
  • the NMS After receiving the NE_STARTUP-REPORT message, the NMS deletes all the entries in the RSJ of the NMS: sends the NEJTARTUP-C0NFI message to the device side; sends ALL_RS_I_S to all the operation status block RS to the device side. – REQ message;
  • the device side After receiving the NE-STARTUP message, the device side resets the ne-start Jioldin g of the corresponding network management (it indicates that the device can process other packets of the NMS and report it); Stop the T-NE_STARTUP of the network management system. REPORT timing, no longer send NE-STARTUP JEP0RT packets to the NMS; start all RS R buffers reporting timers for the NMS:

Landscapes

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

Abstract

本发明涉及一种分组设备运行状态同歩方法,分组设备承载不同类型亚务时,某一种业务的一条业务的运行状态对应一个运行状态块条目RS_I,一个运行状态块条目RS_I的内容包括:条目摘要RS_I_S和条目数据,该某一种业务的所有业务的运行状态块条目RS_I的集合与块的摘要RS_S组成该种业务的运行状态块RS,当网管主动执行全量同歩或设备侧增量上报,或当同步运行状态条目或同步运行状态块时,均采用先同步摘要再同步内容的策略。本发明所述的同步方法,将增量同歩和全量同步相结合,采用把摘要和具体数据相分离的数据组织方法,能很好地完成运行状态同歩,在提高运行状态同歩效率的同时,也能保证运行状态的准确性。

Description

一种分组设备运行状态同步方法
技术领域
本发明涉及分组通信领域中网管和分组设备间运行状态同歩,具 体说是一种分組设备运行状态同步方法。 背景技术
在分組 ( Packet )通信领域,包括二层 VPN (虛拟专用网络》 Virtual Private Network) 、 三层糊、 LSP (标签交換路径, Label Switch Path )等业务由网管部署到设备后, 在设备侧经过控制平面协议动态 分发后才能真正建立相应业务, 当网络变化后, Λ务的状态也会发生 变化, 业务动态建立或者业务状态变化后的运行状态(此状态不是仅 仅包含少量数据的筒单状态, 而是包含大量动态业务信息)需要及时 上报绘网管以保持网管与设备的运行状态的同步 (Synchronous) 。 在网管或者设备重启后, 网管也需要及时获取设备的运行状态, 以保 持同步。
目前,运行状态的同歩都釆用全量同歩的方法, 每次同歩状态的 所有内容。 ¾备数目 (网元数目)多了之后, 运行状态的同歩数掘量 非常巨大, 既效率低下又浪费系统資源,很多时候造成此功能不能使 用。 在大型分组网络中 (网元数目 达几千) , 运行状态同歩的效率 至关重要, 这一问题急需解决。 发嘿 ή容
针对现有技术中存在的缺自,本发明的目的在于提供一种分组设 备运行状态同步方法,将增量同歩 全量同歩相结合, 采用把摘要和 具体数据相分离的数据组织方法, 能很好地完成运行状态同步,在提 高运行状态同步效率的同时, 也能保证运 状态的准确性。 为达到以上目的, 本发明采取的技术方案是:
一种分组设备运行状态同步方法, 其特征在于: 分组设备承载不 同类型业务时,某一种业务的一条业务的运行状态对应一个运行状态 块条目 RS J ,—个运行状态块条目 RS J的内容包播:条目摘要 RS— S 和条目数据,
该某一种业务的所有业务的运行状态块条目 RS— I的集合与块的 摘要 RS— S组成该种 务的运行状态块 RS ,
当网管主动执行全量同步或设备侧增量上报,或当同步运行状态 条目或同歩运行状态块对, 均釆用先同歩摘要再同歩内容的策略, 所述同步摘要指; 同步条目摘要 或同步块的摘要 RS— S ; 所述同步 ή容指: 同步条目数摇, 或同步块的数《, 块的数搦指 某一种业务的所有业务的运行状态块条目 RS— I的集合。
在上述技术方案的基础上, 设备侧增量上报时, 只上报设备侧发 生变化的运行状态条目,先向网管增量上报有变化的运行状态块条目 RS— I的条目摘要 RSJ— S ; 网管收到上报的条目摘要 RS— I— S后, 向设 备侧捞取上报中对应于运行状态块条目 RS Γ的条目数据。
在上述技术方案的基础上, 网管主动执 全量同步时,先同步块 的摘要 RS S和条目摘要 RS— I— S并比较网管侧和设备侧的不同,只同 步网管侧和设备侧的不同部分。
在上述技术方案的基础上, 网管主动执行全量同步的具 ·体步骤 为- 步骤(1 )首先, 网管从设备侧逐个获取运行状态块 RS的块的摘 要 RS— S , 与网管缓存的该运行状态块 RS的块的摘要 KS . 进行比较, 如果一致, 则获取下一个运行状态块 RS的块的摘要 RS— S, 重复上述 比«, 直至全部的运行状态块 RS的块的摘要 均比 «完毕; 如果 不一致, 则执行步骤 (2 ) ;
歩骤 ( 2 ) 从设备侧捞取该运行状态块 RS 所有的条目的摘要 RS^LS , 然后逐条与网管上保存的该运行状态块 RS 的条目的摘要 RS— I— S进行比较, 如果一致, 获取下一条并重复上述比较, 直至全 部条目的摘要 RSJLS均比较完毕; 如果不一致, 则执行步骤 (3); 步骤 (3 ) 从设备侧捞取对应运行状态条目的 RS J [的内容。
本发明所逮的分组设备运行状态同歩方法,将增量同步和全量同 歩相结合,釆用把摘要和具体数据相分离的数据组织方法, 能很好地 完成运行状态同步, 在提高运行状态同步效率的同时, 也能保证运行 状态的准确性。 附圈《«
本发明有如下附圏:
图 1为本发明提供的分组设备运行状态块 RS的组织结构; 图 2为本发明提供的分组设备运行状态块上报缓沖区 RS R (Run
Status Block Report) 的组织结构;
图 3为本发明提供的 W管和设备运行状态组织 ¾构;
圏 4为本发明提供 RSJJJNCILREPORT报文结
图 5为本发明提供的 RS— IJffiCLRUN STATUS报文结构;
图 6为本发明提供的 RS— I— RES报文结构;
图 7为本发明提供的 RSJJiEQ报文结构;
图 8为本发明提供的 RS J€S报文结构;
图 9为本发明提供的 ALL— RS— I— S— 1EQ报文结构;
圏 10为本发明提供的 NE—STARTUPJffiPORT报文结构;
图 11为本发明提供的 NE— STARTUP— REPORT— C0NF0匪报文结构; 图 12为本发明提供的设备运行状态同歩方案的设备倒增量上报 的交互过程图
图 13为本发明提供的设备运行状态同步方案的网管主动执行全 量同歩的交换过程图;
圈 14为本发明提供的设备运行状态同步方案的设备侧重启后, 网管全量同步过程图。 具体实麄方式
以下结合附 a对本发明作进一歩详細说明。
本发明所述的分组设备运行状态同步方法,给出了一种有效的分 组设备运行状态同歩方案,整体上说包括以下四部分: 数据的组织方 法 (摘要和数据相分离)、 先同歩摘要再同歩内容的策略、 备侧增 量上报和网管主动执行全量同步相结合策略。
本发明所述 分组设备运行状态同歩方法, 包括以下歩骤: 步骤 1 , 通过对业务的运行状态内容的组织将其桉摘要和数据两 大类进行分离, 所述按摘要和数据两大类进行分离指: 将用于记载业 务的运行状态的运行状态块 RS划分为运行状态块的摘要 RS_ S (简称 为块的摘要或 与运行状态块的数据 (简称为块的数摇) , 块 的数据为若干运行状态块条目 RS I (簡称为条目 RS I或 RS I或条 目) 的集合, 因此, 再桉摘要和数掘两大类进行分离, 将每个运行状 态块条目 RS— I划分为条 摘要 RSJ— S (简称为 RS_LS)与条目数撫; 参删 1;
分组设备承载 LSP、 二层漏、 三层 VPN等若千不同类型业务, 其中某一种业务的一条业务 运行状态对应一个运行状态块条目 RSJ ( Run Status Item) 。 一个运行状态块条目 RS J的内容包摇: 条目摘要 RS— I— S (Run Status Item Sumiiary, RS— I— S) 和条目数据, 条目摘要 RS_LS和条目数据根据业务的类型而定, 显然不同业务有 不同的条目摘要和条目数据。
该某一种业务的所有业务的运行状态块条目 RS— I的集合作为块 的数据, 其与块的摘要 RS— S ( Run Status BLOCK Summary, RS— S ) 組成该种业务的运行状态块 RS (Run Status BLOCK, RS—B) 。 即: 一个运行状态块 RS的内容包括 块的摘要 RS— S和块的数据, 块的数 据指其该块的全部运行状态块条 RS i的集合。
步骤 2, 采用先同步摘要再同步内容的策略, 即先同步摘要再同 步内容, 不管是网管主动执行全量同步还是设备侧增量上报(增量同 步), 不管是同步运行状态条 还是同步运行状态块, 都是遵循先摘 要后内容的策略。 块 (指运行状态块 RS ) 的同歩的顺序为 块的摘 要 RS S. 块的内容 (即块的数据) 。 条目的同步順序: 条目的摘要
(条目摘要 RSJLS) , 条目的内容 (条目数据) 。
歩骤 3, 设备侧增量上 «,
设备侧增量上报则只上报设备侧发生变化的运行状态条目。即当 设备侧一个运行状态块的部分运行状态块条目 RS— I发生变化时, 先 向网管增量上报有变化的运行状态块条目 RS_I的条目摘要 RS— I— S; 管收到上报的条目摘要 后, 向设备侧捞取上报中对应于运 行状态块条目 RSJ的条目数据。
歩骤 4, 网管主动执行全量同歩,
在网管重启、定时或者设备侧重启的情况下, 网管主动向设备侧 发起全量同步,此处的全量同步不是同步所有运行状态块的所有条目 的内容,而是先同歩块的摘要 RS— S和条目摘要 RS— I—S并比较网管倒 和设备侧的不同,然后只同步网管侧和设备侧的不同部分, 具体同步 过程如下:
歩骠(1 )首先, 网管从设备倒逐个获取运行状态块 RS的块的摘 要 RS一 S, 与网管缓存的该运行状态块 KS的块的摘要 RS一 S进行比较, 如果一致, 则获取下一个运行状态块 RS的块的摘要 RS— S, 重复上述 比较,直至全部的运行状态块 RS的块的摘要 RS—S均比较完毕; 如果 不一致, 则执行歩骤 (2);
步骤 (2 ) 从设备侧捞取该运行状态块 RS 所有的条 的摘要 RS— I— S , 然后逐条与网管上保存的该运行状态块 RS 的条目的摘要 RSJ— S进行比较, 如果一致, 获取下一条并重复上述比较, 直至全 部条目的摘要 RS— I— S均比较完毕; 如果不一致, 则执行歩骤 (3);
歩骤(3 )从设备侧捞取对应运行状态条目 RS— I的内容(该内容 即指条目数据) 。 捞取完一条运行状态条 RS I 的条 数据后, 获 取下一条运行状态条目 RS— I,继续步骤(2 )所述的条目的摘要 RS— I— S 的比较, 同理, 处理完一个运行状态块 RS后, 获取下一个运行状态 块 RS , 继续步骤 ( 1 ) 所述的块的摘要 RS— S的比较。 为了更好地理解本发明的技术方案, 以下通过实施例进一步描 述。
1、 相关数据的组织
( 1 ) 分组设备运行状态块 RS的组织结
图 1是分組设备一种类型运行状态块 HS的組织结构, 一个运行 状态块 RS的内容包含块的摘要 KS_S和块的数据,块的数据指该运行 状态块全部运行状态块条目 RSJ的集合, 其中块的摘要 RS S包插: 类型 BJ>pe、 特征位 BJlag、 条目总数 I—Count 块序列号 B—Seq、 块变化时间戮 B— Sta*p、 块校验和 B— CRC, 块的摘要 RS— S中的 B— Seq 和 BJtamp—起反狭整个运行状态块的变化, 即运行状态块 RS的任 意一个条目的任意变化, Seq加一, 并 i B— Staiip也发生变化。 每 一运行状态块条目 RSJ的内容包含条目摘要 RS— I— S和条目的数据, 条目摘要 包括: 条目标识 1^0)、 条目序列号 ; LSeq、 条目特 征位 I lag 条目校验和] LCRC, 其 Φ Seq反映该条目的变化, 该 条目的数据变化一次, I Seq加一。 条目特征位 I— Flag表示是添加、 删除或更新, 还表示上报是否需要确认。
( 2 ) 运行状态块上报缓冲区的組织结构
图 2 是设备某一种类型运行状态块上报缓冲区 ( Run Status
Block Report , RSJO 的组织结构, 每一个运行状态块上报缓冲区 RS— R, 包含 RS S 指针 (指向运行状态块 RS 中的 RS S 地址)、
B— Seq— Confirmed、 I— Report— Count和运行状态改变需要上报的运行 状态块上报条目摘要 RS J— S。 B— SeiCoirf irmed是网管对上报的最新 一个块 Seq (B— Seq ) 的确认值, 这样网管重新启动后, 可以继续进 行增量同歩; I— Report— Co丽 t是上报条目数。 RS— R中的 RS— I— S是对 应运行状态块 RS的相应条目摘要的拷贝。 设备侧为每一个运行状态 块对应每台网管生成一个 RS— R, RS R 包括需要上报的 RS— I的摘要 d— s。
(3 ) 网管和设备数据的组织
图 3网管和设备以运行状态块为单位的数据組织。
1 )一个运行状态块 RS存放一种类型运行状态块的内容(包括块 的摘要 和块的数摁)
2)运行状态块上报缓冲区 RS— R, 缓冲区内存储的是 RS部分信 息 (部分的意思是只存放需要上报的部分条目的摘要 RS^LS)的拷贝, 缓存发生改变需要上报的条目 RS— I的摘要, 其中包括需要上报值还 未上报的条目和已经上报信还未收到确认的条目。 每当 RS有一个条 目 RS J发生变化, 就在 RS— R查找对应 条目, 如果没有找到 就创建一个新的条目; 如果找到 RSJLS, 就更新该条目。 一个 RSJC 条目的多次变化只在 RS— R占据一个 RS— I— S条目, 不会占掘多条。
3 ) 在设备上对于每一种业务存在一个唯一与之对应的运行状态 块 RS, 所有网管共享一个 RS (针对该种业务), 为每一个网管维持一 个 RS— R„
4) 在网管上为每种运行状态块 RS维持一个运行状态块 RS和一 个运行状态块上报缓冲区 RS— R, 运行状态块的同步就是保持网管与 设备的 RS的同歩。
5) 网管 Φ的 RS R缓存设备 RSJJJNCRJffiPORT报文上报已经 变化的条目的摘要,担 a没有收到该条目的内容;如果没有收到内容, 就永远不删除该条目, 除 Φ收到该条目的最新内容。 如果条目标识
I— ID和条目序列号 I— Seq都一致, 则删除此条 的缓存。 如不同, 比较 I— Seq后, 用大的 I— Seq替代缓冲区对应条目的 I— Seq。
6)设备在一个定时器中上报 RS—R中所有条目。收到网管的确认 后, 从 RS— R删除对应条目 RSJ— S (不是全部删除, 防止在上报和确 认之间有新增加的条目)。 如果 I— ID—样, 就删除此条目, 不用判新 I— Seq是否一致。
7) 设备 RSJJiES报文从 RS中生成。
8) 设备 RS—I—S—INCR— REPORT报文从 RS R中生成。 9) 网管的 RS— I—REQJXMFI腿报文从网管 RS j中生成。
10 )设备有一个全局 start Jiolding标志,控制着设备在重启 期间, 上报的处理。
(4) 下面是实施细节中使用到的报文:
1 )图 4, RSJJJNCILREPORT报文: 设备侧向网管上报运 ί亍状 态块增量变化, 上报变化了条目的摘要。
2)图 5, RSJJ^ CONFI誦报文: 网管向设备侧请求一个运行 状态块的某些条 的内容, 并对上报条 确认。
3) 图 6, ALL— RS— I— S— REQ报文: 网管向设备侧请求一个运行 状态块的所有条目摘要。
4)图 7, RS丄 RES报文: 设备侧向网管回答运行状态块条目的 内容。
5)图 8, 1«^^1¾¾ ¾文: W管向设备侧请求运行状态块的摘要。
6)图 9, RS— S—RES报文: 设备侧向网管回答运行状态块的摘要。
7 ) 圏 10, NE— STARTUP— REPORT报文 设备侧向网管上报重启。
8)图 11» NE— STARTUP JMNFIRMED报文: 网管对设备侧向网管上 报重启 (NE— STARTUP—丽 PORT报文) 的确认。
2、 处理流程
( 1 ) 设备侧运行状态块增量上报的交互流程
如圏 12所示, 具体过程描述如下;
1 ) 当设备侧运行状态块某一个条目内容变化时, 将 RS 的 BJeq++、 BJtamp^当前时间戮; 修改 RS J的摘要 ( I— Seq+等修改) 和数据; 添加或者修改 RS K中的 RS— I— S。
2)当设备侧上报定时器超时,将 RS— R中的 RS— S拷贝到报文 RS— S 字段, 遍历 RS R表中每一 RS— 生成 RSJ— S— INCILREP0RT ¾文 向网管上报。
3) 网管收到 RS—I—S—INCR— REPORT报文, 修改自己 RS R摘要, 其中将报文中的 B— Seq拷贝到 RS R的 SeqJ^nf irmed字段, 添加 或者修改 RS— R中的 RS— I— S。 4) 网管维护 RS R的定时器超时, 遍历 RS R中的 RS— I—S生成
RSJ—REQJDNFI丽报文, 其中将 RS中的 B JeqJ¾nf irmed拷贝报文 B_Seq字段, 向设备侧发送 RS—I—REQ— CONFIRM报文。
5)当设备侧收到 RS— IjEQJXMFI固报文,棍据报文中 Seq字 段修改 RS中的 B Jeq—Conf irmed,遍历报文中的每一个 RS J J条目, 以条目的 ID作为关键字从 RS— R表 φ查找 RS— I— S,如果找到则从 RS— R 表中删除该 RS— 条目 (表示收到上报确认);从 RS表中查找 RS I , 如果找到,在 RS— LIES中为该条目生成一条完整 RS—I信息, 如果没 有找到, 表示该报文被删除。 在 RS— I— RES中只为该条目生成 RS— I— S 摘要, 摘要中的 LFlag 中表示删除的特 ffi位置位; 向网管发送 RS— I— RES报文。
6 )当网管收到 RS— I— RES报文, 遍历报文中的每一个 RS— I条目, 以条目的 ID作为关键字从网管 RS R表中查找 RS I S , 如果找到, 并且接收的 Seq大于 RS R表 Seq, 则删除该 RS— I—S条目 (表示收到 内容, 不需要再捞取条目的内容了); 从 RS表中查找 RS— I, 如果找 到并比较 Seq, 如果报文中的大, 则修改 RS一 I信息; 如果没有找 到, 则新增一条 RS— I; 并修改 RS摘要。
( 2 ) 网管主动执行运行状态块全量同歩流程
如圍 13所示, 具体过程描述如下:
1 ) 网管向设备侧发送一个运行状态块的 RS— SJffiQ报文。
2 ) 设备侧收到 RS— S— REQ报文后, 根据 RS 的摘要向网管发送 RS— S—RES报文 0
3;)网管收到设备侧 RS. S. Ri:S报文后,比较网管 RS和报文的 B— Seq 与 H S t amp , 如果完全相同, 则退出全量同步过程; 否则比较 RS中 的 B—Seq和报文的 Se(LConfirmed,如果相同则退出全量同步过程; 否则将 RS φ现存 RS— I— S 添加到 RS— R 中, 并向设备侧发送 ALL— RS— I— S— REQ报文。
4) 设备侧收到 ALL— RS— IJ— REQ报文后, 将 RS的所有条 RSJ 的摘要拷贝到 RS— R†。 5 ) 设备侧上报定时器超时, 遍历 RS R 中的 RSJLS 生威 RS J J JNCILREP0RT报文, 然后向网管发送该报文。
6)后续的处理参见设备侧运行状态块增量上报的后续处理流程。
(3 ) 设备侧重启后, 网管全量國步运行状态块流程
如 a 14所示, 具体过程描述如下:
1 )设备侧重启, 将每台网管对应的 ne— start—holding置位(在 此位置位期间设备除了处理 NE— STARTUP 画报文外不处理任何 其他报文); 启动 T-NE— STARTUP—REPORT 定时器 (每台网管), 发送 NE— STARTUP—匪 PORT报文给网管;
2)网管收到 NE— STARTUP—REPORT报文后, 删除网管所有 RSJ?中 所有条目: 向设备侧发送 NEJTARTUP— C0NFI画报文; 向设备侧发送 所有运行状态块 RS的 ALL— RS— I— S— REQ报文;
3 ) 设备侧收到 NE— STARTUP 麗 报文后, 对对应网管的 ne— start Jiolding 复位 (表示设备侧可以处理该网管其他报文并上 报); 停 lh该网管的 T-NE— STARTUP— REPORT定时 , 不再向该网管发 送 NE—STARTUP JEP0RT报文; 启动所有 RS R对应该网管的缓沖区上 报定时器:
4 )设备侧收到 ALL— RS— I— S— REQ报文之后的流程, 与网管主动执 行全量同步的后续流程相同。 本说明书中未作详细描途的内容属于本领域专业技术人员公知 的现有技术。

Claims

权 利 要 求 书
1 . 一种分组设备运行状态同步方法, 其特征在于: 分组设备承 载不同类型业务时,某一种业务的一条业务的运行状态对应一个运行 状态块条目 RSJ[, 一个运行状态块条目 RSJ的内容包括: 条目摘要 RS— I— S和条目数据,
该某一种业务的所有业务的运行状态块条目 RS— I的集合与块的 摘要 RS— S组成该种业务的运行状态块 RS,
当网管主动执行全量同歩或设备侧增量上报,或当同步运行状态 条目或同歩运行状态块时, 均采用先同步摘要再同歩内容的策略, 所述同步摘要指; 同步条目摘要 RSJLS,或同步块的摘要 RS— S; 所述同歩内容指: 同歩条目数据, 或同歩块 数据, 块的数据指 某一种业务的所有 Λ务的运行状态块条目 RS— I的集合。
2. 如权利要求 1所述的分组设备运行状态同步方法, 其特征在 于: 设备侧增量上报时, 只上报设备侧发生变化的运行状态条目, 先 向网管增量上报有变化的运行状态块条目 RSJ的条目摘要 RS^LS; 网管收到上报的条目摘要 RS— I— S后, 向设备侧捞取上报中对应于运 行状态块条目 RS— I的条目数据。
3. 如权利要求 1所述的分组设备运行状态同步方法》 其特征在 ]"-: 网管主动执行全量同步时, 先同步块的摘要 RS— S ¾条目摘要 RS— I— S 并比较网管侧和 i殳备侧的不同, 只同步网管侧 设备侧的不 同部分。
4. 如权利要求 3所述的分组设备运行状态同步方法, 其特征在 于, 网管主动执行全量同歩的具体步骤为:
步骤 ( 1 )首先, 网管从设备侧逐个获取运行状态块 RS的块的摘 要 RS一 S, 与网管缓存的该运行状态块 RS的块的摘要 RS一 S进行比较, 如果一 ¾, 则获取下一个运行状态块 RS的块的摘要 RS— S, 重复上述 比较, 直至全部的运行状态块 RS的块的摘要 RS S均比较完毕; 如果 不一致, 则执行歩骤 (2); 步骤 (2) 从设备侧捞取该运行状态块 RS 所有的条目的摘要 RS^LS, 然后逐条与网管上保存的该运行状态块 RS 的条目的摘要 RS— I— S 进行比较, 如果一致, 获取下一条并重复上述比较, 直至全 部条目的摘要 RSJLS均比较完毕; 如果不一致, 则执行步骤 (3);
步骤 (3) 从设备侧捞取对应运行状态条目的 RS—I的内容。
PCT/CN2013/079953 2013-03-20 2013-07-24 一种分组设备运行状态同步方法 WO2014146388A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310090082.9A CN103117883B (zh) 2013-03-20 2013-03-20 一种分组设备运行状态同步方法
CN201310090082.9 2013-03-20

Publications (1)

Publication Number Publication Date
WO2014146388A1 true WO2014146388A1 (zh) 2014-09-25

Family

ID=48416162

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/079953 WO2014146388A1 (zh) 2013-03-20 2013-07-24 一种分组设备运行状态同步方法

Country Status (4)

Country Link
CN (1) CN103117883B (zh)
CL (1) CL2015001013A1 (zh)
MY (1) MY171466A (zh)
WO (1) WO2014146388A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103117883B (zh) * 2013-03-20 2015-09-16 烽火通信科技股份有限公司 一种分组设备运行状态同步方法
CN113726845A (zh) * 2021-07-14 2021-11-30 深圳市有为信息技术发展有限公司 车载终端的数据传输方法、装置、车载终端和商用车辆
CN113938381A (zh) * 2021-10-15 2022-01-14 牙木科技股份有限公司 一种通用简单的网管和网元配置同步的方法和系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1480858A (zh) * 2002-09-03 2004-03-10 鸿富锦精密工业(深圳)有限公司 非结构化数据同步系统及方法
US20050005089A1 (en) * 2003-07-02 2005-01-06 Agere Systems, Inc. Condition management system and a method of operation thereof
CN102292936A (zh) * 2011-05-27 2011-12-21 华为技术有限公司 链路状态数据库同步检测方法及路由设备
CN103117883A (zh) * 2013-03-20 2013-05-22 烽火通信科技股份有限公司 一种分组设备运行状态同步方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3750504B2 (ja) * 2000-08-09 2006-03-01 セイコーエプソン株式会社 データ更新方法および情報処理装置
CN1249597C (zh) * 2002-09-03 2006-04-05 鸿富锦精密工业(深圳)有限公司 分布式文件同步系统及方法
CN101009516B (zh) * 2006-01-26 2011-05-04 华为技术有限公司 一种进行数据同步的方法、系统及装置
CN1874270A (zh) * 2006-06-23 2006-12-06 北京佳讯飞鸿电气有限责任公司 一种实现远程网管的数据同步方法
CN1937536A (zh) * 2006-10-31 2007-03-28 华为技术有限公司 网管拓扑数据同步刷新的方法及系统
CN101436924B (zh) * 2008-11-06 2011-09-14 华为终端有限公司 一种数据同步的方法、装置及系统
CN102054035B (zh) * 2010-12-29 2013-01-02 北京播思软件技术有限公司 一种基于数据范围的数据库数据同步方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1480858A (zh) * 2002-09-03 2004-03-10 鸿富锦精密工业(深圳)有限公司 非结构化数据同步系统及方法
US20050005089A1 (en) * 2003-07-02 2005-01-06 Agere Systems, Inc. Condition management system and a method of operation thereof
CN102292936A (zh) * 2011-05-27 2011-12-21 华为技术有限公司 链路状态数据库同步检测方法及路由设备
CN103117883A (zh) * 2013-03-20 2013-05-22 烽火通信科技股份有限公司 一种分组设备运行状态同步方法

Also Published As

Publication number Publication date
CN103117883A (zh) 2013-05-22
CL2015001013A1 (es) 2015-07-10
MY171466A (en) 2019-10-15
CN103117883B (zh) 2015-09-16

Similar Documents

Publication Publication Date Title
CN108512739B (zh) 以太网虚拟专用网络中的多宿主路由器之间的组播状态
Zhang et al. Scalable name-based data synchronization for named data networking
EP2421206B1 (en) Flooding-based routing protocol having database pruning and rate-controlled state refresh
US8442046B2 (en) Sending routing protocol data on a multi-access network segment
WO2011063757A1 (zh) 一种备份方法、设备和系统
WO2009067865A2 (en) Method, router, line card and active master card for realizng a link state database synchronization
CN106453651B (zh) 一种rpki资料库及数据同步方法
US20140201267A1 (en) Method, apparatus, and system for migrating user service
WO2018149408A1 (en) High availability using multiple network elements
US20140297774A1 (en) System for managing configuration updates in cluster of computational devices
WO2014090083A1 (zh) 分布式弹性网络互连的业务承载方法及装置
US20160080340A1 (en) Communication control device
WO2015154420A1 (zh) 基于isis的路由计算方法及装置
WO2011103759A1 (zh) 关联的双向标签交换路径的创建方法及系统
JP2016502336A (ja) リングネットワークプロテクションにおけるラベル自動割当てのための方法及び装置
EP3157211B1 (en) Isis-based flooding method and device
WO2012129921A1 (zh) 组播方法及组播设备
US10020952B1 (en) PIM relay mode in computer networks
EP3166263B1 (en) Routing calculation method and device for trill isis
WO2014146388A1 (zh) 一种分组设备运行状态同步方法
WO2013083013A1 (zh) 一种网络设备间的同步方法、网络设备及系统
WO2012041142A1 (zh) 一种建立和拆除双向点到多点标签转发路径的方法及系统
WO2014169856A1 (zh) 一种组播通信方法和汇聚交换机
JP5950021B2 (ja) 制御装置、通信システム、同期方法及びプログラム
CN107872387B (zh) 发送vpn路由的方法和系统

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: 13878951

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: 13878951

Country of ref document: EP

Kind code of ref document: A1