Fault tolerant computer system with fault isolation

Claim 2

A fault tolerant computer system comprising: a central processing system including at least one set of data pathways, and means for executing a series of data processing in­structions including the transfer of messages along said plurali­ty of data pathways; at least one set of transaction data storage means, coupled to said at least one set of data pathways, for storing a prede­termined number of successive messages transferred most recently on said at least one set of data pathways; error checking means, for detecting the presence of errors in said central processing system; and error storage means, coupled to said at least one set of transaction data storage means and said error checking means, for causing said at least one set of transaction data storage means to cease storing additional messages in response to the detection of errors by said error checking means.

Claim 3

The fault tolerant computer system of claim 1 further comprising: means for collecting data from said at least one set of transaction data storage means in response to an error detection by said error checking means.

Claim 4

The fault tolerant computer system of claim 2 futher including means for analyzing said collected data to isolate the source of errors in said at least one set of pathways.

Claim 5

The computer system of claim 2 further comprising: means for comparing corresponding depths of stored data within said at least one set of transaction storage means so as to detect whether a data fault of unknown location has passed through said at least one set of data storage means.

Claim 6

The fault tolerant computer system of claim 1 further comprising: means for comparing the data stored at a given level of storage in said at least one set of transaction data storage means; means, responsive to the comparison of stored data in said at least one set of transaction data storage means, for isolating portions of said at least one set of data pathways when the com­parison of stored data at any depth of storage indicates that a fault has not passed through said at least one set of transaction data storage means.

Claim 7

The fault tolerant computer system of claim 5 compris­ing: at least a second set of transaction data storage means lo­cated along said at least one set of data pathways whereby said means for comparing data at a given level of storage in said at least one set of transaction data storage means compares data at the corresponding level of storage at the second set of transac­tion storage means when the data comparison at said first set of transaction data storage means indicates that a fault has not passed therethrough.

Claim 8

The fault tolerant computer system of claim 1 wherein the number of sets of transaction data storage means is suffi­cient to isolate all replaceable portions of said data pathways.

Claim 9

The fault tolerant computer system of claim 6 wherein said means for isolating portions of said at least one set of data pathways is operative to isolate those portions between said first and second sets of transaction data storage means in re­sponse to a comparison of stored data in said second set of transaction data storage means indicating that a fault has not passed through said second set of transaction data.

Claim 10

A fault tolerant computer system comprising: a first central processing system including at least one set of first data pathways, and first means for executing a series of data processing instructions including the transfer of messages along said at least one set of first data pathways; a second central processing system including at least one set of second data pathways, and second means for executing said series of data pro­cessing instructions including the transfer of messages along said at least one set of second data pathways; synchronizing means, coupled to said first and second cen­tral processing systems, for maintaining the execution of said series of data processing instructions by said first and second central processing systems in synchronism; cross-link communications means for transferring messages between said first and second central processing systems via cross-link pathways; transaction data storage means, coupled to said at least one set of first data pathways and to said at least one set of second data pathways and to said cross-link pathways, for storing a pre­determined number of successive messages transferred most re­cently on said sets of first and second data pathways and on said cross-link pathways; error checking means, for detecting the presence of errors in said first and second central processing systems; and error storage means, coupled to said transaction data stor­age means and said error checking means, for causing said trans­action data storage means to cease storing additional messages in response to the detection of errors by said error checking means.

Claim 11

A fault tolerant computer system comprising: a first central processing system including at least one set of first data pathways, and first means for executing a series of data processing instructions including the transfer of messages along said at least one set of first data pathways; a second central processing system including at least one set of second data pathways, and second means for executing said series of data pro­cessing instructions including the transfer of messages along said at least one set of second data pathways; synchronizing means, coupled to said first and second cen­tral processing systems, for maintaining the execution of said series of data processing instructions by said first and second central processing systems in synchronism; cross-link communications means for transferring messages between said first and second central processing systems via cross-link pathways; transaction data storage means, coupled to said at least one set of first data pathways and to said at least one set of second data pathways and to said cross-link pathways, for storing a pre­determined number of successive messages transferred most re­cently on said sets of first and second data pathways and on said cross-link pathways; error checking means, for detecting the presence of errors in said first and second central processing systems; and error storage means, coupled to said transaction data stor­age means and said error checking means, for causing said trans­action data storage means to cease storing additional messages in response to the detection of errors by said error checking means.

Claim 12

The fault tolerant computer system of claim 1 further comprising: means for collecting data from said transaction data storage means in response to an error detection by said error checking means.

Claim 13

The fault tolerant computer system of claim 10 wherein the collected data is sufficient to isolate the source of fault along each of said first and second sets of pathways or said cross-link pathways.

Claim 14

The computer system of claim 10 further comprising: means for comparing corresponding depths of stored data within said transaction data storage means so as to detect wheth­er a data fault of unknown location has passed through said transaction data storage means.

Claim 15

The fault tolerant computer system of claim 9 further comprising: means for comparing the data stored at a given level of storage in said transaction data storage means; means, responsive to the comparison of stored data in said transaction data storage means, for isolating portions of said sets of data pathways when the comparison of stored data at any depth of storage indicates that a fault has not passed through said transaction data storage means.

Claim 16

The fault tolerant computer system of claim 13 compris­ing: at least second transaction data storage means located along said sets of data pathways whereby said means for comparing data given compares data at the same level of storage in the second transaction storage means when the data comparison of said trans­action data storage means indicates that a fault has not passed therethrough.

Claim 17

The fault tolerant computer system of claim 9 wherein the number of transaction data storage means is sufficient to isolate all replaceable portions of said sets of data pathways.

Claim 18

The fault tolerant computer system of claim 14 wherein said means for isolating portions of said sets of data pathways is operative to isolate those portions between said first and second transaction data storage means in response to a comparison of stored data in said second transaction data storage means indicating that a fault has not passed through said second trans­action data storage means.

Claim 19

The fault tolerant computer system of claim 9 wherein said transaction data storage means includes a plurality of cir­cular buffers.

Claim 20

The fault tolerant computer system of claim 9 wherein said error checking means includes a comparator for detecting as an error an inequality in the data on two paired data pathways.

Claim 21

The fault tolerant computer system of claim 9 further including a diagnostic processor, coupled to said transaction data storage means, for gathering the messages stored in said transaction data storage means after the detection of an error by said error checking means.