Automated diagnosis of software crashes

1. A method for diagnosing software crashes, the method comprising:

calculating, by one or more processors, using vector space modeling, angles between vectors representing stack-traces from reports in a repository on one or more server computer systems, to define similarities in the reports, wherein reports are deemed similar where angles between their respective vectors are less than a threshold value;

grouping, by one or more processors, the reports into similar sets using a maximal cliques process; and

automatically diagnosing, by one or more processors, in response to a received software crash, a first stack-trace extracted from a first report of the received software crash, to determine a potential solution to the received software crash.

2. The method of claim 1 , wherein automatically diagnosing in response to the received software crash further comprises selecting the potential solution by searching the grouped reports for a stack-trace similar to the first stack-trace.

3. The method of claim 1 , further comprising:

retrieving, by one or more processors, a stack-trace from a report in the repository on the one or more server computer systems; and

automatically creating, by one or more processors, a vector from the retrieved stack-trace, wherein each function name within the stack-trace is a dimension of the resultant vector, and the vector is associated with the report containing the stack-trace.

4. The method of claim 3 , wherein retrieving a stack-trace from a report in the repository on the one or more server computer systems further comprises:

scanning, by one or more processors, for at least one of a new problem report, updated problem report and authorized analysis report files; and

parsing, by one or more processors, the stack-trace from the at least one of new problem report, updated problem report and authorized analysis report files.

5. The method of claim 1 , wherein calculating, using vector space modeling, angles between the vectors further comprises:

computing, by one or more processors, the angle between each pair of vectors; and

determining, by one or more processors, if the angle between each pair of vectors is less than one degree.

6. The method of claim 1 , wherein grouping the reports into similar sets using a maximal cliques process further comprises:

creating, by one or more processors, an undirected graph wherein a vertex represents each report;

connecting, by one or more processors, a pair of reports, determined to be similar, by an edge;

extracting, by one or more processors, one or more separate connected component subgraphs from the undirected graph;

determining, by one or more processors, a maximal clique for each subgraph, wherein each maximal clique represents a set of similar reports;

grouping, by one or more processors, the reports contained in the maximal clique together; and

storing, by one or more processors, cliques into a repository.

7. The method of claim 1 , wherein automatically diagnosing, in response to the received software crash, the first stack-trace extracted from the first report of the software crash, to determine a potential solution to the received software crash, further comprises:

extracting, by one or more processors, the first stack-trace from the first report;

identifying, by one or more processors, a set of reports from the grouped reports that have a similar stack-trace to the extracted first stack trace; and

selecting, by one or more processors, a potential solution from among program temporary fixes associated with the similar stack trace.

8. A computer program product for diagnosing software crashes, the computer program product comprising:

one or more computer readable storage media and program instructions stored on the one or more computer readable storage media, the program instructions comprising:

program instructions to calculate, using vector space modeling, angles between vectors representing stack-traces from reports in a repository on one or more server computer systems, to define similarities in the reports, wherein reports are deemed similar where angles between their respective vectors are less than a threshold value;

program instructions to group the reports into similar sets using a maximal cliques process; and

program instructions to automatically diagnose, in response to a received software crash, a first stack-trace extracted from a first report of the received software crash, to determine a potential solution to the received software crash.

9. The computer program product of claim 8 , wherein the program instructions to automatically diagnose in response to the received software crash further comprise program instructions to select the potential solution by searching the grouped reports for a stack-trace similar to the first stack-trace.

10. The computer program product of claim 8 , further comprising:

program instructions to retrieve a stack-trace from a report in the repository on the one or more server computer systems; and

program instructions to automatically create a vector from the retrieved stack-trace, wherein each function name within the stack-trace is a dimension of the resultant vector, and the vector is associated with the report containing the stack-trace.

11. The computer program product of claim 10 , wherein the program instructions to retrieve a stack-trace from a report in the repository on the one or more server computer systems further comprise:

program instructions to scan for at least one of a new problem report, updated problem report and authorized analysis report files; and

program instructions to parse the stack-trace from the at least one of new problem report, updated problem report and authorized analysis report files.

12. The computer program product of claim 8 , wherein the program instructions to calculate, using vector space modeling, angles between the vectors further comprise:

program instructions to compute the angle between each pair of vectors; and

program instructions to determine if the angle between each pair of vectors is less than one degree.

13. The computer program product of claim 8 , wherein the program instructions to group the reports into similar sets using a maximal cliques process further comprise:

program instructions to create an undirected graph wherein a vertex represents each report;

program instructions to connect a pair of reports, determined to be similar, by an edge;

program instructions to extract one or more separate connected component subgraphs from the undirected graph;

program instructions to determine a maximal clique for each subgraph, wherein each maximal clique represents a set of similar reports;

program instructions to group the reports contained in the maximal clique together; and

program instructions to store cliques into a repository.

14. The computer program product of claim 8 , wherein the program instructions to automatically diagnose, in response to the received software crash, the first stack-trace extracted from the first report of the software crash, to determine a potential solution to the received software crash, further comprise:

program instructions to extract the first stack-trace from the first report;

program instructions to identify a set of reports from the grouped reports that have a similar stack-trace to the extracted first stack trace; and

program instructions to select a potential solution from among program temporary fixes associated with the similar stack trace.

15. A computer system for diagnosing software crashes, the computer system comprising:

one or more computer processors;

one or more computer readable storage media;

program instructions stored on the computer readable storage media for execution by at least one of the one or more computer processors, the program instructions comprising:

program instructions to calculate, using vector space modeling, angles between vectors representing stack-traces from reports in a repository on one or more server computer systems, to define similarities in the reports, wherein reports are deemed similar where angles between their respective vectors are less than a threshold value;

program instructions to group the reports into similar sets using a maximal cliques process; and

program instructions to automatically diagnose, in response to a received software crash, a first stack-trace extracted from a first report of the received software crash, to determine a potential solution to the received software crash.

16. The computer system of claim 15 , wherein the program instructions to automatically diagnose in response to the received software crash further comprise program instructions to select the potential solution by searching the grouped reports for a stack-trace similar to the first stack-trace.

17. The computer system of claim 15 , further comprising:

program instructions to retrieve a stack-trace from a report in the repository on the one or more server computer systems; and

program instructions to automatically create a vector from the retrieved stack-trace, wherein each function name within the stack-trace is a dimension of the resultant vector, and the vector is associated with the report containing the stack-trace.

18. The computer system of claim 17 , wherein the program instructions to retrieve a stack-trace from a report in the repository on the one or more server computer systems further comprise:

program instructions to scan for at least one of a new problem report, updated problem report and authorized analysis report files; and

program instructions to parse the stack-trace from the at least one of new problem report, updated problem report and authorized analysis report files.

19. The computer system of claim 15 , wherein the program instructions to calculate, using vector space modeling, angles between the vectors further comprise:

program instructions to compute the angle between each pair of vectors; and

program instructions to determine if the angle between each pair of vectors is less than one degree.

20. The computer system of claim 15 , wherein the program instructions to group the reports into similar sets using a maximal cliques process further comprise:

program instructions to create an undirected graph wherein a vertex represents each report;

program instructions to connect a pair of reports, determined to be similar, by an edge;

program instructions to extract one or more separate connected component subgraphs from the undirected graph;

program instructions to determine a maximal clique for each subgraph, wherein each maximal clique represents a set of similar reports;

program instructions to group the reports contained in the maximal clique together; and

program instructions to store cliques into a repository.