Abstract: Methods and systems for ranking analysis results based on user perceived problems of a database system are described. During operation, an embodiment may receive a designation of a perceived database system performance problem from a user, wherein the problem is associated with one or more database system metrics. Next, the embodiment may determine a set of recommendations for alleviating the perceived database system performance problem. The embodiment may then analyze the set of recommendations to determine, for each recommendation in the set of recommendations, an impact the recommendation is expected to have on the one or more database system metrics. Finally, the embodiment may rank the set of recommendations according to the impact each recommendation is expected to have on the perceived database performance problem.

Abstract: Methods and systems for collecting database performance data. During operation a database server can receive a set of T-SQL statements, wherein the database server executes on a computer system and provides access to a database. Next, the database server can execute the set of T-SQL statements, thereby causing the database server to: (1) create an external script which, when executed by the computer system, collects performance data in a file; (2) execute the external script on the computer system; (3) wait for the external script to complete execution; and (4) load the performance data from the file into a table in the database.

Abstract: Methods and systems for adjusting trace filters in a database system are described. During operation, an embodiment may receive, from a user, an indication of a performance problem. Next, the embodiment may determine settings of one or more trace filters based on the indication of the performance problem. The embodiment may then collect trace data based on the one or more trace filters. While the trace data is being collected, the embodiment may perform the following set of operations one or more times: (1) obtain database system performance metrics, (2) analyze the trace data to obtain an analysis result, and (3) adjust the settings of the one or more trace filters based on the database system performance metrics and the analysis result.

Abstract: Methods and systems for identifying one or more indexes are described. A system can collect a set of database operations that were performed during a given time window. Next, the system can partition the set of database operations based on a set of database tables that were accessed by the set of database operations. Then, for each database table in the set of database tables, the system can determine a set of indexes that are candidates for improving database system performance. Next, based on the candidate indexes, the system can identify one or more indexes, one or more combinations of indexes, and/or a consolidated index that improve overall database system performance.

Abstract: Methods and systems for identifying one or more indexes are described. A system can collect a set of database operations that were performed during a given time window. Next, the system can partition the set of database operations based on a set of database tables that were accessed by the set of database operations. Then, for each database table in the set of database tables, the system can determine a set of indexes that are candidates for improving database system performance. Next, based on the candidate indexes, the system can identify one or more indexes, one or more combinations of indexes, and/or a consolidated index that improve overall database system performance.

Abstract: Methods and systems for collecting performance data are described. During operation a database server can receive a set of T-SQL statements, wherein the database server executes on a computer system and provides access to a database. Next, the database server can execute the set of T-SQL statements, thereby causing the database server to: (1) create an external script which, when executed by the computer system, collects performance data in a file; (2) execute the external script on the computer system; (3) wait for the external script to complete execution; and (4) load the performance data from the file into a table in the database.

Abstract: Methods and systems for ranking analysis results based on user perceived problems of a database system are described. During operation, an embodiment may receive a designation of a perceived database system performance problem from a user, wherein the problem is associated with one or more database system metrics. Next, the embodiment may determine a set of recommendations for alleviating the perceived database system performance problem. The embodiment may then analyze the set of recommendations to determine, for each recommendation in the set of recommendations, an impact the recommendation is expected to have on the one or more database system metrics. Finally, the embodiment may rank the set of recommendations according to the impact each recommendation is expected to have on the perceived database performance problem.

Abstract: Methods and systems for adjusting trace filters in a database system are described. During operation, an embodiment may receive, from a user, an indication of a performance problem. Next, the embodiment may determine settings of one or more trace filters based on the indication of the performance problem. The embodiment may then collect trace data based on the one or more trace filters. While the trace data is being collected, the embodiment may perform the following set of operations one or more times: (1) obtain database system performance metrics, (2) analyze the trace data to obtain an analysis result, and (3) adjust the settings of the one or more trace filters based on the database system performance metrics and the analysis result.