Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberCN1653847 B
Publication typeGrant
Application numberCN 03811410
PCT numberPCT/US2003/016449
Publication date8 Dec 2010
Filing date22 May 2003
Priority date24 May 2002
Also published asCA2486800A1, CA2486800C, CN1653847A, DE60323830D1, EP1508259A2, EP1508259B1, US7808924, US7821969, US20030218986, US20080159172, WO2003101144A2, WO2003101144A3
Publication number03811410.0, CN 03811410, CN 1653847 B, CN 1653847B, CN-B-1653847, CN03811410, CN03811410.0, CN1653847 B, CN1653847B, PCT/2003/16449, PCT/US/2003/016449, PCT/US/2003/16449, PCT/US/3/016449, PCT/US/3/16449, PCT/US2003/016449, PCT/US2003/16449, PCT/US2003016449, PCT/US200316449, PCT/US3/016449, PCT/US3/16449, PCT/US3016449, PCT/US316449
Inventors克劳迪奥德桑蒂, 马尔科迪贝内代托
Applicant思科技术公司
Export CitationBiBTeX, EndNote, RefMan
External Links: SIPO, Espacenet
Apparatus and method for preventing disruption of fibre channel fabrics caused by reconfigure fabric (RCF) messages
CN 1653847 B
Abstract
An apparatus and method for preventing the disruption of Fibre Channel Fabrics caused by ReConfigure Fabric (RCF) messages is disclosed. The apparatus includes a storage area network and a plurality of Fibre Channel Switches arranged in a Fabric. Each of the plurality of Switches includes logic to selectively configure their Ports to either reject or accept RCF messages. When configured to rejectRCF messages, the Switch Port that receives an RCF message will generate a reject message along with a reason code explanation ''E_Port Isolated'', and then transition into an Isolated state. When the Switch that generated the RCF message receives the reject message, its Port also transition into the Isolated state. In accordance with the method of the present invention, either a Storage Service Provider or a client can access the Switches of the Fabric through either a command line interpreter or a management application. Once access to the Fabric is established, the logic of the Ports of the Switches can be selectively configured to reject or accept RCF messages as described above.
Claims(28)  translated from Chinese
  1. 一种防止重配置结构消息引起的光纤通道结构扰乱的方法,包括以下步骤:将存储区域网络的结构的光纤通道交换机的一个或多个端口配置为拒绝重配置结构消息,其中,通过将所述交换机的所述一个或多个端口配置为拒绝重配置结构消息,防止所述结构被扰乱;并且使得多个客户能够访问所述结构。 A scrambling method of a Fibre Channel fabric reconfiguration message structure caused prevented, comprising the steps of: a storage area network configuration of the fiber channel switch or more ports configured to reject the reconfiguration message structure, wherein, by the said one or more switch ports configured to reject the reconfiguration message structure, to prevent disruption of the structure; and allows multiple clients to access the structure.
  2. 2.根据权利要求1所述的方法,其中,由存储服务供应商执行所述将一个或多个端口配置为拒绝重配置结构消息的步骤。 2. The method of claim 1, wherein, by the storage service provider to perform the one or more ports configured to reject the message re-configuration of the step structure.
  3. 3.根据权利要求1所述的方法,其中,由客户执行所述将一个或多个端口配置为拒绝重配置结构消息的步骤。 3. The method of claim 1, wherein, by the client to execute the one or more ports configured to reject the message re-configuration of the step structure.
  4. 4.根据权利要求1所述的方法,其中,通过将所述交换机的所述一个或多个端口配置为拒绝所述重配置结构消息,所述结构的所述交换机能够拒绝重配置结构消息。 4. The method of claim 1, wherein said switch by one or more ports configured to reject the message re-arrangement structure, the structure of the switch reconfiguration configuration message can be rejected.
  5. 5.根据权利要求1所述的方法,其中,所述结构的所述交换机的所述一个或多个端口被配置为当接收到重配置结构消息时,生成带有原因代码说明“E端口被隔离”的拒绝消息。 E port 5. The method of claim 1, wherein the configuration of said switch of said one or more ports are configured to when receiving reconfiguration message structure, generated with a reason code explanation "is isolation, "the rejection message.
  6. 6.根据权利要求1所述的方法,还包括通过命令行解释程序或管理应用程序来配置所述结构的所述交换机。 6. The method according to claim 1, further comprising configuring the switch structure by the command line interpreter or management application.
  7. 7.根据权利要求6所述的方法,还包括使用管理动作以将所述结构的所述交换机的所述一个或多个端口配置为拒绝重配置结构消息。 7. The method according to claim 6, further comprising using a management action to the structure of the switch of the one or more ports configured to reject the reconfiguration message structure.
  8. 8.根据权利要求1所述的方法,还包括通过显式管理动作,将所述存储区域网络的所述结构的所述交换机的所述一个或多个端口重新配置为接受重配置结构消息。 8. The method according to claim 1, further comprising through explicit management action, the structure of the switch of the storage area of the one or more network ports reconfiguration message to accept the reconfigurable fabric.
  9. 9.根据权利要求1所述的方法,其中,所述交换机依靠光纤通道协议来进行通信,所述光纤通道协议包括因特网协议上的光纤通道协议以及任何其他技术上的光纤通道协议。 9. The method of claim 1, wherein, said switches rely on the Fibre Channel protocol to communicate, the Fibre Channel protocol includes Fibre Channel Protocol over Internet Protocol, and any other technical Fibre Channel protocol.
  10. 10. 一种防止重配置结构消息引起的光纤通道结构扰乱的方法,包括以下步骤:在存储区域网络的结构处,支持多个交换机;传送来自以及去向使用所述结构的多个客户的通信;以及禁止所述客户在执行调用重配置结构消息的动作时扰乱所述结构的能力,其中,禁止所述客户扰乱所述结构的能力的步骤包括:将所述结构的所述交换机的端口配置为拒绝重配置结构消息。 10. A method of preventing reconfiguration message Fibre Channel fabric structure caused by disturbances, comprising the steps of: at the structure in the storage area network, to support a plurality of switches; and transmitting a communication from a plurality of clients using the whereabouts of the structure; and inhibits the ability to disrupt the structure of the client when the call reconfiguration configuration message actions in execution, wherein the step of inhibiting the ability of the client to disturb the structure comprises: the port configuration of the switch is the structure refuse reconfiguration structure of the message.
  11. 11.根据权利要求10所述的方法,其中,所述结构依靠光纤通道协议来进行通信,所述光纤通道协议包括因特网协议上的光纤通道协议以及任何其他技术上的光纤通道协议。 11. The method of claim 10, wherein said structure to rely on the Fibre Channel protocol for communication, the Fibre Channel protocol includes Fibre Channel Protocol over Internet Protocol, and any other technical Fibre Channel protocol.
  12. 12.根据权利要求11所述的方法,还包括从客户交换机逻辑隔离接收到重配置结构消息的E端口。 12. The method according to claim 11, further comprising receiving from the client to switch the logical isolation structure reconfiguration message E port.
  13. 13.根据权利要求11所述的方法,其中,所述结构的所述交换机的所述端口被配置为当接收到重配置结构消息时,生成带有原因代码说明“E端口被隔离”的拒绝消息。 13. The method according to claim 11, wherein said port of said switch of said structure is configured to, when the received reconfiguration message structure, a reason code explanation "E port is separated" refuse to produce with message.
  14. 14. 一种用于配置被布置在存储区域网络的结构中的多个光纤通道交换机的方法,包括以下步骤:在存储区域网络的结构的光纤通道交换机处,允许管理员建立到所述结构的连接,所述结构包括多个交换机;以及将所述交换机的一个或多个端口配置为拒绝重配置结构消息,其中,通过将所述交换机的所述一个或多个端口配置为拒绝重配置结构消息,防止在重配置结构消息发生时其自身被扰乱。 The method of the plurality of fiber channel switch 14. A configuration is disposed in the configuration of the storage area network, comprising the steps of: a Fibre Channel storage area network at the switch structure, allows an administrator to establish the structure connection, the structure comprises a plurality of switches; and one or more ports of the switch is configured to deny reconfiguration message structure, wherein the switch by the one or more ports configured to reject the reconfigurable fabric message, to prevent the structure of the message occurs when reconfiguring itself is disturbed.
  15. 15.根据权利要求14所述的方法,还包括通过将所述一个或多个交换机的所述端口配置为拒绝所述重配置结构消息,将所述一个或多个交换机配置为拒绝重配置结构消息。 15. The method according to claim 14, further comprising the one or more through the switch port is configured to reject the message re-arrangement structure, the one or more switches configured to reject the reconfigurable fabric message.
  16. 16.根据权利要求15所述的方法,还包括将所述一个或多个交换机的所述端口配置为当接收到重配置结构消息时,生成带有原因代码说明“E端口被隔离”的拒绝消息。 16. The method of claim 15, further comprising one or more switches when the port is configured to receive the reconfiguration configuration message, a reason code explanation "E port is separated" refuse to produce with message.
  17. 17.根据权利要求14所述的方法,还包括通过命令行解释程序或管理应用程序建立到所述结构的所述连接。 17. The method according to claim 14, further comprising a line interpreter or management application to establish the structure of the connection command.
  18. 18.根据权利要求14所述的方法,还包括通过显式管理动作,将所述结构的所述交换机的所述端口重新配置为接受重配置结构消息。 18. The method according to claim 14, further comprising a through explicit management action, the port of the switch structure of the reconfiguration message to accept the reconfigurable fabric.
  19. 19. 一种防止重配置结构消息引起的光纤通道结构扰乱的装置,包括:带有光纤通道结构的存储区域网络;和布置在所述结构中的多个交换机,所述多个交换机的各个端口包括:用于选择性地拒绝重配置结构消息的逻辑,其中,所述交换机的所述逻辑还被配置为当所述逻辑被选择为拒绝重配置结构消息时,充分防止所述结构被重配置结构消息扰乱。 19. A method of preventing reconfiguration message Fibre Channel fabric structure caused by disruption of apparatus, comprising: a storage area network with a Fibre Channel fabric; in the structure and arrangement of a plurality of switches, all ports of the plurality of switches comprising: means for selective rejection of the reconfiguration of the logical structure of the message, wherein the logic is further configured to switch when the logic is selected to reject the reconfiguration message structure, said structure being sufficiently prevented reconfiguration disrupt the structure of the message.
  20. 20.根据权利要求19所述的装置,其中,各个交换机的所述逻辑使得各个交换机的E端口当所述逻辑被选择为拒绝重配置结构消息时,拒绝重配置结构消息。 20. The apparatus of claim 19, wherein the logic is such that each switch port E each switch when the logic is selected to reject the reconfiguration message structure, message reject the reconfigurable fabric.
  21. 21.根据权利要求20所述的装置,其中,各个交换机的所述逻辑使得各个交换机的E端口当接收到重配置结构消息并且所述逻辑被选择为拒绝重配置结构消息时,生成带有原因代码说明“E端口被隔离”的拒绝消息。 21. The apparatus of claim 20, wherein each of said logic port E of each switch such that the switch structure when the received reconfiguration message and the logic is selected to reject the reconfiguration message structure, generated with the reasons Code Description "E port is isolated," the rejection message.
  22. 22.根据权利要求19所述的装置,其中,各个交换机的所述逻辑还被配置为选择性地允许各个E端口分别接受重配置结构消息。 22. The apparatus according to claim 19, wherein each switch of the logic is further configured to selectively allow each E-port structure respectively accept reconfiguration message.
  23. 23.根据权利要求19所述的装置,还包括用于提供到所述结构的访问的连接,所述连接提供从命令行解释程序或管理应用程序的访问。 23. The apparatus according to claim 19, further comprising a connection for providing access to the structure, the connection provides a command line interpreter or management application access.
  24. 24.根据权利要求19所述的装置,其中,所述结构依靠光纤通道协议来进行通信,所述光纤通道协议包括因特网协议上的光纤通道协议以及任何其他技术上的光纤通道协议。 24. The apparatus of claim 19, wherein said structure to rely on the Fibre Channel protocol for communication, the Fibre Channel protocol includes Fibre Channel Protocol over Internet Protocol, and any other technical Fibre Channel protocol.
  25. 25. 一种被配置为被包括在存储区域网络的结构中的交换机,所述交换机包括:交换机端口;以及用于选择性地拒绝重配置结构消息的逻辑,其中,所述交换机的所述逻辑还被配置为当所述逻辑被选择为拒绝重配置结构消息时,充分防止所述结构被重配置结构消息扰乱。 25. A switch configured to be included in the structure of the storage area network, said switch comprising: a switch port; and means for selectively rejecting reconfiguration logic structure of the message, wherein the logic switch When further configured to be selected when the reconfigurable logic architecture news for refusing to fully prevent the structure of the message structure is disrupted reconfiguration.
  26. 26.根据权利要求25所述的交换机,其中,所述逻辑被配置为选择性地使得所述交换机的E端口当所述逻辑被选择为拒绝重配置结构消息时,拒绝重配置结构消息。 26. The switch of claim 25, wherein the logic is configured to selectively such that the port E of the switch when the logic is selected to reject the reconfiguration message structure, message reject the reconfigurable fabric.
  27. 27.根据权利要求26所述的交换机,其中,所述逻辑还被配置为如果接收到重配置结构消息并且所述逻辑被选择为拒绝重配置结构消息,则使得E端口生成带有原因代码说明“E端口被隔离”的拒绝消息。 27. The switch of claim 26, wherein, wherein the logic is further configured to, if the received reconfiguration message and the structure of the logic is selected to reject the reconfiguration message structure is such that the port E is generated with a reason code explanation "E port is isolated," the rejection message.
  28. 28.根据权利要求25所述的交换机,其中,所述逻辑可以被配置为允许重配置结构消息被接受。 28. A switch according to claim 25, wherein, the logic may be configured to allow reconfiguration configuration message is accepted.
Description  translated from Chinese

防止重配置结构消息引起的光纤通道结构扰乱的装置和方 Preventing the Fibre Channel fabric reconfiguration message structure caused disruption apparatus and

France

技术领域 FIELD

[0001] 本发明涉及存储区域网络,更具体地,涉及用于防止由重配置结构(RCF)消息引起的光纤通道结构(Fabric)扰乱的装置和方法。 [0001] The present invention relates to storage area networks, and more particularly, relates to a structure for preventing the reconfiguration (RCF) message due to a Fibre Channel fabric (Fabric) apparatus and method for disruption.

背景技术 BACKGROUND

[0002] 随着因特网商务和以网络为中心的计算的日益普及,商业和其他组织正变得越来越依赖信息。 [0002] With the increasing popularity of Internet commerce and network-centric computing, commercial and other organizations are becoming increasingly dependent on information. 为了处理所有这些数据,存储区域网络或SAN已经变得非常普遍。 In order to deal with all these data, storage area network, or SAN has become very common. SAN通常包括许多存储设备、许多主机以及布置在交换结构中的多个交换机。 SAN typically includes many storage devices, many hosts and arranged in a plurality of switch fabric switches. 交换机在SAN内选择性地连接存储设备和主机。 Switches selectively connect storage devices and hosts in the SAN.

[0003] 大多数SAN依靠光纤通道协议以在结构内通信。 [0003] Most SAN Fibre Channel protocol relies on communication within the structure. 关于光纤通道协议和光纤通道交换结构的详细解释,见Fibre Channel Framing and SignalingStandard(光纤通道成中贞禾口信令标准),Rev. L 70, American NationalStandard of Accredited Standards Committee(NCITS),2002 年2 月8 日和Fibre Channel Switch Fabric (光纤通道交换结构)-2,Rev. 5.4,NCTIS,2001年6月26日,这里为了适于各种用途,通过引用结合了这两篇文献。 For a detailed explanation of the Fibre Channel protocol and Fibre Channel switch fabric, see Fibre Channel Framing and SignalingStandard (Fibre Channel into the mouth Wo Ching signaling standard), Rev. L 70, American NationalStandard of Accredited Standards Committee (NCITS), 2002 February March 8 and Fibre Channel Switch Fabric (Fibre Channel switch fabric) -2, Rev. 5.4, NCTIS, 2001 年 6 26, here is suitable for a variety of purposes, through a combination of these two literature references.

[0004] 在光纤通道中,用由制造商指定的唯一的八⑶字节宽节点名(Node_Name)标识每个设备(主机、存储设备和交换机)。 [0004] In Fibre Channel, with only eight ⑶ byte-wide node name (Node_Name) specified by the manufacturer to identify each device (host, storage devices, and switches). 当光纤通道设备被互联以形成SAN时,节点名(连同其他参数)被用来标识每个设备。 When the Fibre Channel devices are interconnected to form a SAN, the node name (along with other parameters) is used to identify each device. 光纤通道帧被用于在SAN中的设备之间通信。 Fibre Channel frames are used for communication between the SAN devices. 但是, 帧不使用节点名。 However, the frame does not use the node name. 而是,每个终端设备(主机和存储设备)的光纤通道端口通过由结构动态分配给终端设备的三(3)字节光纤通道地址(或FC_ID)而被寻址。 Instead, each end device (hosts and storage devices) of Fibre Channel ports through dynamically allocated to the terminal apparatus by a structure of three (3) byte Fibre Channel address (or FC_ID) which is addressed. 每个终端设备通过执行向交换结构的结构登录过程获得其FC_ID。 Each terminal device obtains its FC_ID login process to the structure by performing the switching fabric. 在该过程中,终端设备与结构交换通过SAN 成功通信所需的操作参数和它们的证书。 In this process, the terminal device operating parameters and structure required for the exchange of successful communication via SAN and their certificates. 最初,结构用唯一的结构名(Fabric^Name)来标识其自身,用终端设备的唯一节点名来标识终端设备。 Initially, to identify its own structure with a unique structure name (Fabric ^ Name), with a unique node name to identify the terminal device terminals. 此后,结构向终端设备指定FC_ID。 Thereafter, the structure specified FC_ID to the terminal device.

[0005] 3字节宽的光纤通道地址被分级构建在3个字段中,每个字段1字节长:域ID (Domain_ID)、区域ID (Area_ID)和端口ID (Port_ID)。 [0005] 3-byte wide Fibre Channel addresses are hierarchically-structured in three fields, each field is 1 byte long: Domain ID (Domain_ID), the area ID (Area_ID) and port ID (Port_ID). 结构内的每个交换机被指定一个域ID。 Each switch is assigned a structure within the domain ID. 连接到具体交换机的终端设备被指定该交换机的域ID。 A switch connected to the specific terminal device is assigned a domain ID of the switch. 交换机管理每个终端设备的区域ID和端口ID字段的分配,以保证在该域中被指定地址的唯一性。 Assigned to each terminal device switch management area ID and Port ID field to ensure the address is specified in the domain of uniqueness. 例如,如果交换机被指定的域号是5,该交换机将其地址空间再分为两个区域,每个区域有三个连接的终端设备,则可能的光纤通道地址分配是:5 : 1 : 1,5 : 1 : 2,5 : 1 : 3,5 : 2 : 1,5 : 2 : 2 和5 : 2 : 3。 For example, if the switch is specified domain number is 5, the switch to its address space is further divided into two regions, each with three terminal device is connected, you may Fibre Channel address allocation is: 5: 1: 1, 5: 1: 2,5: 1: 3,5: 2: 1,5: 2: 2 and 5: 2: 3.

[0006] 当交换结构初始化时,交换机中的一个被选为主交换机。 [0006] When the switch fabric initialization, the switch in a selected master switch. 主交换机向结构中所有交换机指定域ID,其节点名变为交换结构的结构名。 The master switch specifies the domain ID to all the switches in the structure, and its name changed to the structure of the node name of the switch fabric. 为了选择主交换机,所有交换机都彼此交换被称为交换结构参数(EFP)的消息。 To select the main switch, all switches are called swap exchanging structural parameters (EFP) messages. 除了其他参数外,EFP包括发送交换机的节点名。 In addition to other parameters outside, EFP including sending switch node name. 具有最低节点名的交换机被指派为主交换机。 Switch with the lowest node name is assigned the master switch. 所有其他交换机被称为非主交换机。 All other switches are called non-main switch. 一旦选定了主交换机,主交换机就向其邻近交换机发送域标识符指定(DIA)消息,该消息通知邻 Once you have selected a master switch, the master switch on their proximity switch sends the specified domain identifier (DIA) message that notifies neighborhood

4近交换机其已经被指定了主交换机域ID。 4 near the switch that it has been designated a master switch domain ID. 作为回复,邻近交换机向主交换机发送请求域标识符(RDI)消息。 As a response, the neighboring switch sends a request to the master switch domain identifier (RDI) message. 主交换机分配域ID,并通过向每个交换机发送其域ID作出响应。 The master switch allocation domain ID, and respond by sending its domain ID to each switch. 此后, 接收到域ID的交换机向它们的邻近交换机发送DIA,接收回复中的RDI,并向为发出请求的交换机指定域ID的主交换机转发该RDI。 Thereafter, the receiver switches to the domain ID is sent to their proximity switch DIA, receive a reply in RDI, RDI and forwards the request to the master switch switches the specified domain ID. 该过程持续直至所有交换机接收到域ID。 This process continues until all the switch receives a domain ID. 在已经接收到域ID后,各个交换机对其域中的每个终端设备指定区域ID和端口ID。 After having received the domain ID, and each switch port ID specified area ID of each terminal device to its domain. 当所有交换机已经被分配了域ID时,认为结构配置完成。 When all switches have been assigned domain ID, consider the structure of the configuration is complete. 由此终端设备都被指定了它们的区域ID 和端口ID。 Thus terminal devices are assigned to their area ID and port ID.

[0007] 光纤通道使得两个分立交换结构能合并为一个。 [0007] Fibre Channel switch fabric so that the two separate energy into one. 这发生于在两个分属于不同结构的交换机之间建立连接时。 This occurs when the connection is established between the two switches belonging to different structures. 当发生这样的事件时,被合并的结构的某些交换机的域ID和它们的终端设备的FC_ID可能需要重新指定。 When such event, FC_ID domain ID Some switches are merged structure and their terminal equipment may need to be specified. 例如,如果结构A将要与第二结构B合并,其中结构A包括域ID —、二和三(1、2和3),结构B包括域ID —和二(1和2),则两个结构中一个结构的重叠的域ID(1和2)必须重新指定。 For example, if the structure of A will be merged with the second structure B, where A structure including the domain ID -, II and III (1, 2 and 3), the structure B including domain ID - and two (1 and 2), the two structures overlapping domain ID in a structure (1 and 2) must be re-specified. 当两个结构连接时,通过连接它们的链路,交换EFP消息,以判断域ID中是否有重叠。 When the two structures connected by connecting their link, exchange EFP message, the domain ID to determine whether there is overlap. 基于结果,可能发生两种情况中的一种。 Based on the results, in both cases one may occur.

[0008] 如果交换机之间的域ID指定有重叠,则连接两个初始结构的链路被隔离。 [0008] If the domain ID of the switch between the designated overlap, the two initial structures connecting link is isolated. 虽然物理链路还存在,但是该链路被逻辑断开,且不能被设备识别。 Although the physical link is still present, but the link is a logical disconnect device and can not be identified. SAN管理员然后可以请求对连接结构的扰乱性重配置,以解决域ID冲突。 SAN administrator can then request a connection structure of disruptive reconfiguration to resolve domain ID conflicts. 在这种情况中,重配置结构(RCF)消息泛滥通过两个初始结构的所有交换机。 In this case, the reconfigurable fabric (RCF) message flooding by all switches two initial structure. 这停止了所有数据流量的传送,并且每个交换机取消其域ID。 This stops the transmission of all the data traffic, and each switch to cancel the domain ID. 此后,以与上述方式相同的方式选定主交换机,向交换机指定新的域ID,向终端设备指定新的FC_ID。 Thereafter, in the same manner as described above selected master switch, specify the new domain ID to the switch, specify a new FC_ID to the terminal device. 以这种方式,两个结构被合并为一个。 In this manner, two structures were merged into one.

[0009] 如果在交换机之间没有域ID重叠,则自动执行非扰乱性结构重配置。 [0009] If no domain ID overlap, the automatic execution of non-disruptive structural reconfiguration between switches. 建立结构(BF)消息泛滥通过两个初始结构的所有交换机。 Establish the structure (BF) message flooding through all switches two initial structures. 数据帧传送没有停止,每个交换机保持其域ID。 Data frame transmission is not stopped, each switch maintains its domain ID. 因为两个结构中的每个都有主交换机,所以两个主交换机中的一个必须“辞去”主地位,只剩下一个主交换机用于合并结构。 Because the two structures each with master switch, so the two main switches in a must "resign" Lord position, only a master switch for the combined structure. 因此,发生上述主交换机选择过程。 Therefore, the occurrence of the main switch selection process. 然后每一非主交换机向续存的主交换机发出RDI请求,请求其在BF消息前拥有的相同的域ID。 Then each non-master switch to the surviving master switch sends RDI request, the request the same domain ID in their possession before the BF message. 以这种方式,两个结构被合并,而没有改变任何交换机域ID指定或任何指定给终端设备的FC_ID。 In this manner, the two structures are merged without changing any switch or any specified domain ID assigned to the terminal device FC_ID.

[0010] 但是,对于两个初始结构中的一个结构,主交换机改变了。 [0010] However, for the two initial structures of a structure, the main switch has changed. 因此,这些交换机的结构名必须更新。 Therefore, the structure of the name of these switches must be updated. 假定结构名是每个终端设备维护的结构登录状态信息的一部分,则释放的结构的交换机必须重新初始化它们的终端设备,以更新它们的新状态。 Structure name is part of the structure is assumed login status information maintained by each terminal device, the configuration of the release switch must re-initialize their end devices to update their new status. 该过程导致在失去主交换机的结构中的数据流量的扰乱。 This process leads to the loss of the main switch to disrupt the structure of the data traffic.

[0011] 从SAN可用性的角度,RCF和BF消息两者都是潜在地存在问题的。 [0011] From the perspective of SAN availability, both RCF and BF messages are potentially problematic. 在两个不同结构被合并的情况中,BF可能导致数据流量的部分扰乱。 Two different structures are merged in the case of, BF may lead to some disruption of data traffic. 通常,如果在单个结构内部调用BF 协议,则不会发生扰乱。 Normally, if you call the BF protocol in a single internal structure, it will not disturb occur. 但是,在主交换机被标识,域ID和FC_ID被重新指定时,RCF总是会停止通过结构的数据流量。 However, the main switch is identified, the domain ID and FC_ID be re-specified, RCF will always stop by the data flow structure. 结果,RCF消息引起了严重的安全问题,因为结构中的交换机可能用于发起RCF,这会扰乱该结构的功能。 The results, RCF news caused serious security problems, because the structure of the switch may be used to initiate RCF, which would disrupt the function of the structure.

[0012] 传统上使用“物理安全”解决该安全担忧,其中结构的所有交换机都包含在安全位置内,只有系统管理员或其他被授权的人员才被允许访问该结构。 [0012] use the traditional "physical security" to resolve the security concerns, including the structure of all the switches are included in the safety position, only the system administrator or other authorized personnel are allowed access to the structure. 以这种方式,诸如用户或客户之类的第三方就不能访问该结构的交换机以及触发RCF。 In this manner, a user or a third party such as a client or the like can not access the configuration switches and trigger RCF. 但是只要交换机没有被限制在安全位置中,例如在多建筑的企业园区或者存储服务供应商(SSP)等情形中,这种方法就不是令人满意的。 But as long as the switch is not restricted to a secure location, such as in a multi-building corporate campus or storage service provider (SSP) and other circumstances, this approach is not satisfactory.

5[0013] 对存储服务供应商而言,因为它们通过其基础设施向不同客户提供存储服务,所以该问题尤其严重。 5 [0013] The storage service providers, because they provide storage services to different customers through its infrastructure, so that the problem is especially severe. 如果多个用户被连接到采用分区技术来隔离其流量的同一结构,则一个客户可以触发RCF。 If multiple users are connected to the use of zoning to isolate the flow of the same structure, then a client can trigger RCF. 结果,该结构所服务的所有客户的数据流操作将被扰乱。 As a result, all customers served by the structure of the data stream operations will be disrupted. 从存储服务供应商和客户的角度,无疑都不希望出现这种情景。 From the storage service providers and the customer's perspective, no doubt do not want that scene.

[0014] 为了解决这种担忧,SSP通常向每个客户提供分立的结构。 [0014] To address this concern, SSP typically provide a separate structure to each customer. 这样,如果一个客户希望合并其结构或者以别的方式更新其结构,就不会影响其他客户的结构。 Thus, if a customer wants to consolidate their structure or otherwise update its structure, it will not affect other customers of the structure. 但是,这种布置决不是理想的。 However, this arrangement is not ideal. 对每个客户建立并维护分立结构可能是昂贵的,并且难以管理和支持。 Establish and maintain a separate structure for each client can be expensive and difficult to manage and support.

[0015] 因此,需要一种用于防止由重配置结构(RCF)消息导致的光纤信道结构扰乱的装置和方法。 [0015] Accordingly, a need for an apparatus and method for preventing Fibre Channel structure disturbed by the reconfigurable fabric (RCF) message caused.

发明内容 SUMMARY

[0016] 为了达到上述目的,根据本发明的目的,公开了一种用于防止由重配置结构(RCF) 消息引起的光纤通道结构扰乱的装置和方法。 [0016] In order to achieve the above object, according to the object of the present invention, discloses an apparatus and method for preventing a Fibre Channel fabric by the reconfigurable fabric (RCF) message due to disturbances. 该装置包括存储区域网络和布置在结构中的多个光纤通道交换机。 The apparatus comprises a storage area network and arranged in the structure more Fibre Channel switches. 多个交换机中的每个包括用于选择性地将它们的端口配置为拒绝或者接受RCF消息的逻辑。 Each includes a plurality of switches for selectively configure their ports to reject or accept the logic RCF message. 当被配置为拒绝RCF消息时,接收RCF消息的交换机端口将生成带有原因代码说明“E端口被隔离”的拒绝消息,然后转变为隔离状态。 When configured to reject RCF message, the message is received RCF switch port will generate a rejection message with a reason code explanation "E port is isolated," and then into isolation. 当生成RCF消息的交换机接收到拒绝消息时,其端口也转变为隔离状态。 RCF message is generated when the switch receives a rejection message, the port is also transformed into isolation. 根据本发明的方法,存储服务供应商或客户可以通过命令行解释程序或管理应用程序访问结构的交换机。 The method according to the present invention, the storage service providers or customers through the command interpreter or application access management structure of the switch line. 一旦建立了到结构的访问,交换机的端口的逻辑可以如上所述被选择性地配置为拒绝或接受RCF消息。 Once the access to the structure, logic switch ports as described above can be selectively configured to reject or accept the RCF message.

附图说明 Brief Description

[0017] 图1是根据本发明的存储区域网络的交换结构。 [0017] FIG. 1 is a storage area network switch fabric according to the present invention.

[0018] 图2是根据本发明图示了结构的交换机端口如何被配置以拒绝RCF消息的流程图。 [0018] FIG. 2 is a diagram illustrating how the configuration of the switch ports are configured to reject RCF message flow diagram according to the present invention.

[0019] 图3图示了根据本发明的在核心交换机和客户交换机之间的逻辑断开。 [0019] Figure 3 illustrates the disconnect switch and according to the customer switches between the core logic of the present invention. 具体实施方式 DETAILED DESCRIPTION

[0020] 在下面的描述中,为了提供对本发明的彻底理解,提出了许多具体细节。 [0020] In the following description, in order to provide a thorough understanding of the present invention, numerous specific details. 但是,对本领域技术人员而言清楚的是,没有这些具体细节中的一部分或全部,也可以实施本发明。 However, the skilled artisan is clear that without these specific details of part or all of, the present invention may be practiced. 在其他情况中,没有详细描述公知的处理步骤,以避免不必要地模糊本发明。 In other instances, no detailed descriptions of well-known processing steps, in order to avoid unnecessarily obscuring the present invention.

[0021] 参考图1,示出了可以被存储服务供应商使用的典型存储区域网络的交换结构。 [0021] Referring to Figure 1, there is shown a switch fabric can be used by service providers stored typical storage area network. 存储区域网络(SAN) 10包括交换结构12,该交换结构12包括位于SSP驻地的多个光纤通道交换机SW以及位于客户驻地中的多个客户交换机14。 Storage Area Network (SAN) 10 includes a switch fabric 12, the switch fabric 12 includes SSP premises located more Fibre Channel switches SW and customer premises located in multiple customer switches 14. 每个客户包括一个或多个主机H,主机H通过SSP的结构12和其客户交换机14可以访问多个存储设备D。 Each client includes one or more host H, the host H 14 can access multiple storage devices 12 through the SSP's structure and its customers switch D. 在一个实施例中, 诸如命令行解释程序或管理应用程序16之类的管理结构12的工具通过SSP驻地内的主机18被连接到结构12。 In one embodiment, such as a command line interpreter or management structure management tool application 16 or the like through the host 12 inside the SSP premises 18 is connected to the structure 12. 在另一实施例中,命令行解释程序或管理应用程序16通过连接到客户交换机14的主机H中的一个访问结构12。 In another embodiment, the command line interpreter or management application 16 via the switch is connected to the host client H 14 in an access structure 12. 命令行解释程序或管理应用程序16使得网络管理员能够访问并实施跨越结构12的管理变更。 Command line interpreter or management application enables network administrators to access 16 and implement change management across the structure 12.

[0022] 根据本发明的各种实施例,主机H可以是任何类型的主机,包括但不限于在Unix、Windows或任何其他计算平台上运行的服务器或个人计算机,或者它们的组合。 [0022] In accordance with various embodiments of the present invention, the host computer H may be any type of host, including but not limited to servers or personal computers running on a Unix, Windows, or any other computing platform, or a combination thereof. 类似地, 存储设备D可以是任何类型的存储设备,包括但不限于磁带备份系统、仿真磁带备份系统、 CD-ROM存储阵列或者诸如独立磁盘冗余阵列(RAID)之类的一个或多个盘,或者它们的组合。 Similarly, the storage device D may be any type of storage device, including but not limited to, tape backup systems, tape backup systems simulation, a CD-ROM or a storage array such as a redundant array of independent disks (RAID) or the like or a plurality of disks , or a combination thereof. 交换机SW可以是任何类型的光纤通道交换机,例如可以从加利福尼亚州San Jose的Brocade公司或本发明的受让人Andiamo Systems公司购得的那些交换机。 SW switch can be any type of Fibre Channel switches, such as those switches from Brocade in San Jose, California, company or assignee of the present invention, the company purchased Andiamo Systems. 还应该注意,图中所示的结构12只是对描述本发明有用的SAN的示例。 It should also be noted that the structure shown in Figure 12 is useful for the description of the present invention the SAN. 它的简单性决不能被理解为限制本发明,本发明可以用在任何SAN配置中。 Its simplicity in no way be construed as limiting the present invention, the present invention can be used in any SAN configuration. 例如,本发明例如可以用在有许多客户(多于所图示的两个)的SAN中或用在有多个区域的SAN中。 For example, the present invention may be used for example in many customers (more than the two illustrated) of a SAN or a plurality of regions in the SAN.

[0023] 本发明提供了配置结构12的交换机SW的方式,以选择性地拒绝在它们的端口上接收到的RCF消息。 [0023] The present invention provides a configuration fabric switches SW 12 ways to selectively reject in their port receives the RCF message. 当结构12被初始配置时,或者当在结构12运行后,网络管理员使用命令行解释程序或管理应用程序16想实现跨越结构12的变更时,可以使用该技术。 Time when the structure 12 is the initial configuration, or when in the structure 12 runs, the network administrator uses the command-line interpreter or manage applications across 16 want to achieve structural change 12, you can use the technology. 另外,可以由存储服务供应商通过由该供应商控制的客户或通过被连接到结构12的第三方客户, 通常在存储服务供应商允许下实现该技术,或者直接由SSP自己实现该技术。 Further, by storing the service provider by the customer or the vendor control by being connected to the structure 12 of the third party client, the technique is usually implemented in the permission storage service providers, or directly by the SSP own implementation of this technique.

[0024] 参考图2,根据本发明示出了流程图20,该流程图20图示了结构12的交换机端口如何被配置以拒绝RCF消息。 [0024] Referring to Figure 2, according to the present invention shows a flow chart 20, the flowchart 20 illustrates how the structure of the switch ports 12 are configured to reject RCF messages. 根据FC-SW-2术语系统,术语“端口”被用来指示一般的交换机端口(可能被连接到主机或存储设备或连接到其他交换机),而术语“E端口,,只被用来表示被连接到另一交换机端口的交换机端口。最初,系统管理员通过命令行解释程序或管理应用程序16访问结构12 (框22)。此后,管理员选择结构12中的一个交换机(框24)。 一般地,交换机被选中是因为其可能有接收RCF消息的风险。使用特定的配置命令将被选中交换机的相关端口配置为不接受RCF消息(框26)。配置命令是系统管理员使用命令行解释程序或管理应用程序16向被选中交换机发出的管理指令。配置命令不是“交换机到交换机”消息。其因此没有被FC-SW-2定义。而是,配置命令是交换机的选项,当该选项被设置时,将导致交换机E端口拒绝RCF消息。因此,交换机的行为保持符合FC-SW-2标准。在操作期间,当在任何这样被配置的E端口接收到RCF消息时,E端口生成带有原因代码说明"E端口被隔离”的拒绝消息,此后E端口转变为隔离状态。响应于该拒绝消息,生成该RCF 的交换机的E端口也转变为隔离状态。在菱形判断框28中,管理员判断另一交换机的其他端口是否要被配置为不接受RCF消息。如果是“否”,则序列完成。如果是“是”,则选中另一交换机,重复上述序列。 According to FC-SW-2 terminology, the term "port" is used to indicate a general switch ports (which may be connected to the host or storage device, or connected to other switches), and the term "E port ,, only be used to represent connected to another switch port switch port. Initially, the system administrator via the command line interpreter or management applications access structures 16 12 (block 22). Since then, the administrator selects the structure of a switch 12 (block 24). General , the switch was selected because it may be risky to receive RCF message using a specific configuration commands will be selected for the relevant port configuration of the switch does not accept RCF message (block 26). configuration commands are a system administrator using the command-line interpreter or management instructions issued by the management application 16 is selected to the switch. Configuration command is not "switch-to-switch" message. Thus it is not FC-SW-2 is defined. Instead, the switch configuration command is an option, when the option is set , it will cause the switch port E reject RCF messages. Thus, the behavior of the switch is kept in line with FC-SW-2 standard. During operation, when any E port is configured such RCF message is received, E port generated with cause Code Description "E port is isolated," the rejection message, then E port into isolation. In response to the rejection message, generate E-port switch is also the RCF into isolation in the diamond decision block 28, the administrator judge whether the other port on the other switch is to be configured to not accept RCF messages. If "NO", the sequence is completed. If "YES", then select another switch, repeat the above sequence.

[0025] 参考图3,示出了图示客户交换机和SSP交换机的E端口之间的逻辑断开的图。 [0025] Referring to Figure 3, there is shown a logic diagram illustrating the client switches off the port E and the switch between the SSP. 该图示出了属于SSP的结构12的交换机32的E端口与客户交换机34的E端口被连接36中的中断在逻辑上断开。 The figure shows the structure of the switch 32 to the SSP 12 and customer switch port E E port 34 is connected to the 36 interrupt logically disconnected. 应该注意,断开是逻辑断开,而不是物理断开,如同在光纤通道标准FC-SW-2中描述的当由链路连接的两个E端口处于隔离状态时的结果状态。 It should be noted, is a logical disconnect disconnected, rather than physical disconnection, when the results of the state consists of two E-port link connection in a state of isolation as described in the Fibre Channel standard FC-SW-2 in. 以上述方式,可以配置结构12,使得交换机的端口中的一部分或全部可以被配置为拒绝RCF消息。 In the above-described manner, the structure 12 may be configured such that the switch port a part or all of which may be configured to reject RCF messages. 从实践的观点看,这种安排使得结构12能够被两个(或更多)客户使用。 From a practical point of view, this arrangement enables the structure 12 can be two (or more) clients. 该安排还防止了任何一个客户在试图执行调用RCF消息的动作时扰乱结构12而有害于其他客户。 Disrupt the structure when the arrangement also prevents any attempt to execute a customer's call to action 12 RCF message and harmful to other customers. 结构将保持这种状态,并由此可以被两个或更多客户访问,直至交换机的端口被重新配置为接受RCF消息。 Structure will remain in this state, and thus may be two or more clients access until the switch port is re-configured to accept RCF messages. 通常,只能通过意欲取消端口的前述状态的显式管理动作,才能将交换机的端口重新配置为接受RCF消息。 Typically, only by desire to cancel the previous state of the port of explicit management action to reconfigure the switch port to accept RCF messages.

[0026] 上述的本发明实施例应被认为是示例性的,而不是限制性的。 [0026] The embodiments of the present invention described above should be considered illustrative, and not restrictive. 例如,本发明不是必须与SAN—起使用的。 For example, the present invention is not to be used starting with SAN-. 而是,其可以被如下的任何类型网络使用:该网络具有如光纤通道标准FC-SW-2中描述的那样工作以执行结构配置的光纤通道交换机。 Instead, it can be any type of network used as follows: The network has a work such as the Fibre Channel standard FC-SW-2 to perform as described structural arrangement of the Fibre Channel switches. 本发明并不限于这里给出的细节,而是可以在所附权利要求的范围和等同物内作出修改。 The present invention is not limited to the details given herein, but may be modified within the scope of the appended claims and equivalents thereof.

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
CN1285128A15 Dec 199821 Feb 2001艾利森电话股份有限公司Self-configuring processors in asynchronous transfer mode switch
EP1124172A25 Feb 200116 Aug 2001Emc CorporationControlling access to a storage device
US618520318 Feb 19976 Feb 2001Vixel CorporationFibre channel switching fabric
Non-Patent Citations
Reference
1FIBRE CHANNEL SWITCH FABRIC-2(FC-SW-2)REV 4.0.1998,1-116.
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
CN103597790A *30 May 201219 Feb 2014国际商业机器公司Fibre channel forwarder fabric login sequence
CN103597790B *30 May 20128 Jun 2016国际商业机器公司光纤通道交换机环境中的方法、系统和装置
Classifications
International ClassificationG06F13/10, G06F3/06, H04Q11/00, H04L12/24, H04L29/06
Cooperative ClassificationH04Q11/0005, H04Q2011/0039
European ClassificationH04Q11/00P2
Legal Events
DateCodeEventDescription
10 Aug 2005C06Publication
5 Oct 2005C10Request of examination as to substance
8 Dec 2010C14Granted