Fault tolerant shared system resource with communications passthrough providing high availability communications

• US 6,865,157 B1
• Filed: 05/26/2000
• Issued: 03/08/2005
• Est. Priority Date: 05/26/2000
• Status: Expired due to Term

• Current Assignee: EMC Corporation (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Jones, James Gregory, Scott, John A.
• Primary Examiner(s): Pham, Chi
• Assistant Examiner(s): Ly, Anh Vu H
• Application Number:
  US09/579,428
• Time in Patent Office: 1,747 Days
• Field of Search: 370/217, 370/221, 370/242, 370/245, 370/412, 370/429, 714/1, 714/2, 714/6, 714/10, 714/11, 714/15, 714/25
• US Class Current: 370/242
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/2005 using redundant communication controllers
• CPC Class Codes: G06F 11/2007 using redundant communication media
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups
• CPC Class Codes: H04L 41/00 Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks

Abstract

A communications passthrough mechanism for high availability network communications between a shared system resource and clients of the system resource. The system resource includes a control/processing sub-system including multiple peer blade processors. A port of each blade processor is connected to each client/server network path and each client is connected to a corresponding port of each blade processor. Each blade processor includes a network fault detector exchanging beacon transmissions with other blade processors through corresponding blade processor ports and network paths. Each blade processor includes response generator responsive to a failure to receive a beacon transmission from a failed port of an other blade processor for redirecting the client communications to the failed port on the other blade processor to the corresponding port of the blade processor. A path manager in the blade processor is responsive to operation of the response generator for modifying the communications routing table to correspond with the redirection message to route the client communications to the failed port of the other blade processor to the other blade processor through the inter-processor communications link. Each blade processor may also include an inter-blade communications monitor for detecting a failure in the inter-processor communications link between the blade processor and another blade processor, reading the communications routing table to select a functional network communications path to a port of the other blade processor, and modifying the communications routing table to redirect inter-processor communications to the selected functional network communications path.

Claims

1. A system resource providing system resource services to clients communicating with the system resource through a network including a plurality of client/server communication paths, comprising:

a system resource sub-system for performing system resource operations, and

a control/processing sub-system including

a multiplicity of peer blade processors, each blade processor having a port connected to each client/server network communication path and each client being connected to correspond ports of each of the blade processors, each blade processor including

a network mechanism supporting communications operations between the blade processor and the clients, including a communications routing table identifying the communication routing for each client,

an inter-processor communications processor providing communications between the blade processor and the system resource sub-system and an inter-processor communications link between the blade processors, and

a communications monitoring mechanism, including

a network fault detector for exchanging beacon transmissions with another blade processor through the network communications paths connecting corresponding ports of the blade processors,

a response generator responsive to the network fault detector upon failing to receive a beacon transmission from a failed port of an other blade processor for transmitting a redirection message to the clients redirecting the client communications to the failed port to the corresponding port of the blade processor, and

a path manager responsive to operation of the response generator for modifying the communications routing table to correspond with the redirection message to route the client communications with the other processor blade through the inter-processor communications link.

2. The system resource of claim 1, wherein each blade processor further includes:

an inter-blade communications monitor for

detecting a failure in the inter-processor communications link with another blade processor,

reading the communications routing table to select a functional network communications path between corresponding ports of the blade processor and the other blade processor, and

modifying the communications routing table to redirect inter-processor communications through the inter-processor communications link through the selected functional network communications path.

3. A communications passthrough mechanism for use in a fault tolerant shared system resource for high availability communications between the system resource and clients communicating with the system resource through a network including a plurality of client/server communication paths, comprising:

a system resource sub-system for performing system resource operations, and

a control/processing sub-system including a multiplicity of peer blade processors, each blade processor having a port connected to each client/server network communication path and each client being connected to correspond ports of each of the blade processors,

each blade processor including

a network mechanism supporting communications operations between the blade processor and the clients, including a communications routing table identifying the communication routing for each client,

an inter-processor communications processor providing communications between the blade processor and the system resource sub-system and an inter-processor communications link between the blade processor and another blade processor, and

a communications monitoring mechanism, including

a network fault detector for exchanging beacon transmissions between the blade processor and another blade processor through the network communications paths connecting corresponding ports of the blade processors,

a response generator responsive to the network fault detector upon failing to receive a beacon transmission from a failed port of the other blade processor for transmitting a redirection message to the clients redirecting the client communications to the failed port to the corresponding port of the blade processor, and

a path manager responsive to operation of the response generator for modifying the communications routing table to correspond with the redirection message to route the client communications with the other blade processor through the inter-processor communications link.

4. The system resource of claim 3, wherein each blade processor further includes:

an inter-blade communications monitor for

detecting a failure in the inter-processor communications link between the blade processor and another blade processor,

reading the communications routing table to select a functional network communications path between ports of the blade processor and the other blade processor, and

modifying the communications routing table to redirect inter-processor communications through the selected functional network communications path.

5. A fault tolerant networked file server including communications passthrough mechanism for high availability communications between the file server and clients of the file server through a network including a plurality of client/server communication paths, comprising:

a storage sub-system for storing client file system shares, and

a control/processing sub-system including a multiplicity of peer blade processors, each blade processor having a port connected to each client/server network communication path and each client being connected to correspond ports of each of the blade processors,

each blade processor including

a network mechanism supporting communications operations between the blade processor and the clients, including a communications routing table identifying the communication routing for each client,

an inter-processor communications processor providing communications between the blade processor and the storage sub-system and an inter-processor communications link between the blade processors, and

a communications monitoring mechanism, including

an inter-processor communications processor providing communications between the blade processor and the system resource sub-system and an inter-processor communications link between the blade processor and another blade processor, and

a communications monitoring mechanism, including

a network fault detector for exchanging beacon transmissions between the blade processor and another blade processor through the network communications paths connecting corresponding ports of the blade processors,

a response generator responsive to the network fault detector upon failing to receive a beacon transmission from a failed port of the other blade processor for transmitting a redirection message to the clients redirecting the client communications to the failed port to the corresponding port of the blade processor, and

a path manager responsive to operation of the response generator for modifying the communications routing table to correspond with the redirection message to route the client communications with the other blade processor through the inter-processor communications link.

6. The file server of claim 5, wherein each blade processor further includes:

an inter-blade communications monitor for

detecting a failure in the inter-processor communications link between the blade processor and another blade processor,

reading the communications routing table to select a functional network communications path between ports of the blade processors, and

modifying the communications routing table to redirect inter-processor communications through the selected functional network communications path.

7. In a system resource providing system resource services to clients communicating with the system resource through a network including a plurality of client/server communication paths, a method for providing high availability communications between the system resource and the clients of the system resource, the system resource including a system resource sub-system for performing system resource operations and a control/processing sub-system including a multiplicity of peer blade processors, each blade processor having a port connected to each client/server network communication path and each client being connected to correspond ports of each of the blade processors, each blade processor including a network mechanism supporting communications operations between the blade processor and the clients and an inter-processor communications processor providing communications between the blade processor and the system resource sub-system, comprising the steps of:

in a blade processor,

monitoring communications operations of another blade processor by exchanging beacon transmissions with the other blade processor through the network communications paths connecting corresponding ports of the blade processors,

upon failing to receive a beacon transmission from a failed port of the other blade processor, transmitting a redirection message to the clients redirecting the client communications to the failed port to a corresponding port of the blade processor, and

routing the redirected client communications with the other blade processor through the inter-processor communications link.

8. The method for providing high availability communications between the system resource and the clients of the system resource of claim 7, further comprising the steps of:

in the blade processor,

detecting a failure in the inter-processor communications link between the blade processor and another blade processor,

selecting a functional network communications path between corresponding ports of the blade processor and the other blade processor, and

redirecting inter-processor communications through the selected functional network communications path.