Multiplex communications – Diagnostic testing – Fault detection
Reexamination Certificate
1998-12-30
2003-04-08
Vu, Huy D. (Department: 2665)
Multiplex communications
Diagnostic testing
Fault detection
C370S248000, C370S392000, C714S048000
Reexamination Certificate
active
06545981
ABSTRACT:
BACKGROUND OF THE INVENTION
Traditional network systems utilize either channel semantics (send/receive) or memory semantics (DMA) model. Channel semantics tend to be used in I/O environments and memory semantics tend to be used in processor environments.
In a channel semantics model, the sender does not know where data is to be stored, it just puts the data on the channel. On the sending side, the sending process specifies the memory regions that contain the data to be sent. On the receiving side, the receiving process specifies the memory regions where the data will be stored.
In the memory semantics model, the sender directs the data to a particular location in the memory, utilizing remote direct memory access (RDMA) transactions. The initiator of the data transfer specifies both the source buffer and destination buffer of the data transfer. There are two types of RDMA operations, read and write.
The virtual interface architecture (VIA) has been jointly developed by a number of computer and software companies. VIA provides consumer processes with a protected, directly accessible interface to network hardware, termed a virtual interface. VIA is especially designed to provide low latency message communication over a system area network (SAN) to facilitate multi-processing utilizing clusters of processors.
A SAN is used to interconnect nodes within a distributed computer system, such as a cluster. The SAN is a type of network that provides high bandwidth, low latency communication with a very low error rate. SANs often utilize fault-tolerant capability.
It is important for the SAN to provide high reliability and high-bandwidth, low latency communication to fulfill the goals of the VIA. Further, it is important for the SAN to be able to recover from errors and continue to operate in the event of equipment failures. Error recovery must be accomplished without high CPU overhead associated with all transactions. Furthermore, error recovery should not increase the complexity for the consumer of VIA services.
SUMMARY OF THE INVENTION
According to one aspect of the present invention, a SAN maintains local copies of a sequence number for each data transfer transaction at the requester and responder nodes. Each data transfer is implemented by the SAN as a sequence of request/response packet pairs. An error condition arises if a response to any request packet is not received at the requesting node. The responder and requestor nodes are coupled by a plurality of paths and each node maintains a record of the good or bad status of each path. If a transaction fails and the path is permanently bad, both nodes update their status to indicate that the path is bad. This is to prevent further transactions from including any stale requests that are potentially still in the network from arriving at the destination and potentially corrupting data.
According to another aspect of the invention, if an error occurs on a path the requestor node implements a barrier transaction on the path to determine if the failure is permanent or transient.
According to another aspect of the invention, the barrier transaction is performed by writing a number chosen from a large number space in a way that minimizes the probability of reusing the number in a short period of time.
According to one aspect of the invention, the number is randomly chosen from a large number space.
According to another aspect of the invention, the large number is based on the requestor ID and an incrementing component managed by the requestor.
According to another aspect of the invention, if the failure is transient the requestor retransmits packets starting with the packet that first caused an error condition to be detected.
According to another aspect of the invention, a sequence number is included in each request packet and copied into each response packet. A local copy of the sequence number is maintained at the requestor and responder nodes. If the sequence number in the request packet does not match the sequence number at the responder, a negative acknowledge response packet is generated.
REFERENCES:
patent: 4733350 (1988-03-01), Tone et al.
patent: 5193149 (1993-03-01), Awiszio et al.
patent: 5386524 (1995-01-01), Lary et al.
patent: 5555405 (1996-09-01), Griesmer et al.
patent: 5563879 (1996-10-01), Sanders et al.
patent: 5574849 (1996-11-01), Sonnier et al.
patent: 5619274 (1997-04-01), Roop et al.
patent: 5675579 (1997-10-01), Watson et al.
patent: 5678007 (1997-10-01), Hurvig
patent: 5706281 (1998-01-01), Hashimoto et al.
patent: 5737595 (1998-04-01), Cohen et al.
patent: 5802050 (1998-09-01), Petersen et al.
patent: 5920886 (1999-07-01), Feldmeier
patent: 5991797 (1999-11-01), Furtral et al.
patent: 6119244 (2000-09-01), Schoenthal et al.
patent: 6272591 (2001-08-01), Grun
patent: 6347337 (2002-02-01), Shah et al.
patent: 2002/0029305 (2002-03-01), Satran et al.
patent: 0 757 318 (1997-02-01), None
Garcia et al., “ServerNet II”Parallel Computer Routing and Communication, (2ndInt. WKSP), Jun. 26, 1997, pp. 119-135, XP002103164, Atlanta, GA.
Eicken Von T. et al., “U-Net: A User-Level Network Interrface for Parallel and Distributed Computing,”Operating Systems Review(SIGOPS), vol. 29, No. 5, Dec. 1, 1995, pp. 40-53.
Dunning D. et al., “The Virtual Interface Architecture,”IEEE Micro, vol. 18, No. 2, 3/98, pp. 66-76.
Garcia David J.
Larson Richard O.
Low Stephen G.
Watson William J.
Compaq Computer Corporation
Ho Duc
Oppenheimer Wolff & Donnelly LLP
Sherry Leah
Vu Huy D.
LandOfFree
System and method for implementing error detection and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for implementing error detection and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for implementing error detection and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3036299