Distributed fabric management protocol

• US 9,088,477 B2
• Filed: 02/02/2012
• Issued: 07/21/2015
• Est. Priority Date: 02/02/2012
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Anantharam, Sushma, Ghosh, Nirapada, Kamble, Keshav Govind, Leu, Dar-Ren, Mendon, Chandarani J., Pandey, Vijoy A., Peethambaram, Nandakumar
• Primary Examiner(s): Sefcheck, Gregory
• Assistant Examiner(s): ULYSSE, JAEL M
• Application Number:
  US13/364,896
• Publication Number:
  US 20130201873A1
• Time in Patent Office: 1,265 Days
• Field of Search: 370/219, 370/254, 370/255, 370/396, 370/469, 370/216, 370/316, 370/350, 370/390, 370/401, 709/208, 709/253, 709/206, 709/223
• US Class Current: 1/1
• CPC Class Codes: H04L 41/0246 Exchanging or transporting network management information using the Internet; Embedding network management web servers in network elements; Web-services-based protocols
• CPC Class Codes: H04L 41/044 comprising hierarchical management structures

Abstract

A distributed fabric system comprises a plurality of independent network elements interconnected by inter-switch links and assigned to a same group. Each network element includes a switching chip, a processor, and memory storing program code that is executed by the processor. The program code of each network element includes a device configuration (DC) stacking module and a switch discovery protocol (SDP) module. The SDP module of each network element, when executed, discovers each other network element in the group and elects one of the network elements as a master network element. The SDP module of the master network element, when executed, sends messages to the DC-stacking module of the master network element. Each sent message identifies one of the network elements in the group. The DC stacking module of the master network element, when executed, maintains a record of all network elements that are currently members in the group.

Claims

1. A distributed fabric system comprising a plurality of independent network elements interconnected by inter-switch links (ISLs) and assigned to a same group, each network element including one or more switching chips, a processor, and memory storing layered program code that is executed by the processor, the layered program code of each network element including a device configuration (DC) stacking module and a switch discovery protocol (SDP) module that includes a member tracking layer (MTL) component, the SDP module of each network element, when executed, discovering each other network element in the group, the MTL component of each network element electing one of the network elements as a master network element, the SDP module of the master network element, when executed, sending messages to the DC stacking module of the master network element, each sent message identifying one of the network elements in the group, the DC-stacking module of the master network element, when executed, maintaining a record of all network elements that are currently members in the group, wherein:

the master network element receives a switch discovery protocol data unit (SDPDU) from a remote network element that is not a member of the group and sends, in response to receiving the SDPDU from the remote network element, a message to the remote network element by which the master network element identifies itself to the remote network element as master of the group;

the remote network element, after receiving the message from the master network element by which master network element identifies itself, determines whether the remote network element is ready to join the group; and

the MTL component of the master network element maintains switch information about all network elements that are currently members in the group learned over the ISLs, employs a timer to age out the remote network element if no SDPDU is received from the remote network element for a specified duration, clears switch information learned over a given ISL when the given ISL goes down, and sends a message to the DC-stacking module of the master network element to notify of any change in membership in the group resulting from a newly discovered network element or from a departure of a network element.

2. The distributed fabric system of claim 1, wherein the layered program code of each network element further comprises application programs at a top layer of the layered program code and a DC-API (application program interface) module at a layer beneath the top layer, the DC-API module of each network element being configured to obtain information from and to set parameters of the one or more switching chips of any one of the network elements.

3. The distributed fabric system of claim 2, wherein the layered program code of each network element further comprises a Remote Procedure Call (RPC) module configured to be called by the DC-API module of that network element to communicate with one of the other network elements.

4. The distributed fabric system of claim 1, wherein the layered program code of each network element further comprises an Ethernet based L2 Transport (EL2T) to facilitate communication.

5. The distributed fabric system of claim 4, wherein the layered program code of each network element further comprises a path selection layer (PSL) at a layer below the SDP module and the EL2T layer, the PSL being configured to determine an egress port for a packet when transmitted over an inter-switch link to a remote network element.

6. The distributed fabric system of claim 4, wherein the layered program code of each network element further comprises a trivial file transfer protocol (TFTP) at a layer above the EL2T layer, the master network element using the TFTP to push a configuration and a firmware image to the other network elements.

7. The distributed fabric system of claim 1, wherein the layered program code of each network element further comprises a check-point module (CP) to synchronize applications on the master network element with related databases and states at a backup network element in preparation for a backup-to-master failover.

8. The distributed fabric system of claim 1, wherein the layered program code of each network element further comprises applications at a top layer of the layered program code and a switch attach/detach module configured to provide the applications with a global view of all data ports on all the network elements in the group and to notify the applications of any change on any of the network elements.

9. The distributed fabric system of claim 1, wherein the layered program code of each network element further comprises applications at a top layer of the layered program code and a port mapping/macros module configured to provide the applications with a mapping of a global CLI (command line interface) port to a physical device and a port.

10. The distributed fabric system of claim 1, further comprising sending, by the master network element, a remote procedure call to the remote network element to set a parameter of a hardware switching chip of the remote network element.

11. The distributed fabric system of claim 1, further comprising sending, by the master network element, a remote procedure call to the remote network element to configure a port of a hardware switching chip of the remote network element.

12. The distributed fabric system of claim 1, further comprising sending, by the master network element, a remote procedure call to the remote network element to request information from a hardware switching chip of the remote network element.

13. The distributed fabric system of claim 1, wherein the remote network element determines whether the remote network element is ready to join the group by determining that the remote network element has a same configuration as the master network element.

14. The distributed fabric system of claim 1, wherein the remote network element determines whether the remote network element is ready to join the group by:

requesting a configuration script from the master network element in response to the message by which the master network element identifies itself to the remote network element as master of the group; and

executing the configuration script.

15. A computer program product for managing a distributed fabric system in which a plurality of independent network elements are interconnected by inter-switch links (ISLs), the network elements being assigned to a same group, the computer program product comprising:

a non-transitory computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code comprising;

switch discovery protocol (SDP) module configured to discover each network element in the group, the SDP module including a member tracking layer (MTL) component configured to elect one of the network elements as a master network element, the MTL component of the master network element maintaining switch information about all network elements that are currently members in the group learned over the ISLs and clearing switch information learned over a given ISL when the given ISL goes down, the SDP module, when executed by the master network element, being further configured to issue a message for each network element in the group; and

a device configuration (DC) stacking module configured to communicate switch information with each other network element in the distributed fabric system, the DC-stacking module, when executed by the master network element, being configured to receive each message issued by the SDP module running on the master network element and to maintain a record of all network elements that are currently members in the group,

wherein;

the master network element is configured to receive a switch discovery protocol data unit (SDPDU) from a remote network element that is not a member of the group and to send, in response to receiving the SDPDU from the remote network element, a message to the remote network element by which the master network element identifies itself to the remote network element as master of the group;

the remote network element is configured to, after receiving the message from the master network element by which master network element identifies itself, determine whether the remote network element is ready to join the group; and

the MTL component of the master network element employs a timer to age out the remote network element if no SDPDU is received from the remote network element for a specified duration and sends a message to the DC-stacking module of the master network element to notify of any change in membership in the group resulting from a newly discovered network element or from a departure of a network element.

16. The computer program product of claim 15, wherein the computer-readable program code further comprises application programs at a top layer of a software stack and a DC-API (application program interface) module at a layer beneath the top layer, the DC-API module being configured to obtain information from and to set parameters of the one or more switching chips of any one of the network elements.

17. The computer program product of claim 16, wherein the computer-readable program code further comprises a Remote Procedure Call (RPC) module configured to be called by the DC-API module to communicate with a remote network element.

18. The computer program product of claim 15, wherein the computer-readable program code further comprises an Ethernet based L2 Transport (EL2T) to facilitate communication with a remote network element.

19. The computer program product of claim 18, wherein the computer-readable program code further comprises a path selection layer (PSL) at a layer in a software stack below the SDP module and the EL2T layer, the PSL being configured to determine an egress port for a packet when transmitted over an inter-switch link to a remote network element.

20. The computer program product of claim 18, wherein the computer-readable program code further comprises a trivial file transfer protocol (TFTP) at a layer in a software stack above the EL2T layer, the TFTP being configured to push a configuration and a firmware image to a remote network element.

21. The computer program product of claim 15, wherein the computer-readable program code further comprises a check-point module (CP) configured to synchronize applications on the master network element with related databases and states at a backup network element in preparation for a backup-to-master failover.

22. The computer program product of claim 15, wherein the computer-readable program code further comprises application programs at a top layer of a software stack and a switch attach/detach module configured to provide the applications with a global view of all data ports on all the network elements in the group and to notify the applications of any change on any of the network elements.

23. The computer program product of claim 15, the computer-readable program code further comprises applications at a top layer of a software stack and a port mapping/macros module configured to provide the applications with a mapping of a global CLI (command line interface) port to a physical device and a port.