Log SQL statements

Administrators can optionally log all SQL statements used in sql macros. This option is enabled by adding configuration in the ../WEB-INF/classes/log4j.properties file. In addition, warnings are issued when the default row limit is exceeded. Timing logging is also available since version 5.1.0.

Add to ../WEB-INF/classes/log4j.properties

# SQL statement logging
log4j.appender.sqlplugin=org.apache.log4j.RollingFileAppender
log4j.appender.sqlplugin.File=${catalina.home}/home/logs/sql-plugin.log
log4j.appender.sqlplugin.Threshold=INFO
log4j.appender.sqlplugin.MaxFileSize=20480KB
log4j.appender.sqlplugin.MaxBackupIndex=10
log4j.appender.sqlplugin.layout=com.atlassian.confluence.util.PatternLayoutWithContext
log4j.appender.sqlplugin.layout.ConversionPattern=%d [%t] %m%n

log4j.logger.org.swift.confluence.sql.SqlMacro.statement=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlMacro.statement=false

log4j.logger.org.swift.confluence.sql.SqlQueryMacro.statement=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlQueryMacro.statement=false

log4j.logger.org.swift.confluence.sql.SqlFileMacro.statement=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlFileMacro.statement=false
 
# SQL time logging
log4j.logger.org.swift.confluence.sql.SqlMacro.timing=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlMacro.timing=false

log4j.logger.org.swift.confluence.sql.SqlQueryMacro.timing=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlQueryMacro.timing=false
 
log4j.logger.org.swift.confluence.sql.SqlFileMacro.timing=INFO, sqlplugin
log4j.additivity.org.swift.confluence.sql.SqlFileMacro.timing=false

Add logging at runtime

Go to Administration -> Logging and Profiling
Find the Add new entry area
Add INFO entries for the following: org.swift.confluence.sql.SqlMacro.statement and org.swift.confluence.sql.SqlQueryMacro.statement
Log statements will appear in the standard confluence log file
Logging will occur immediately, but only last until Confluence is restarted

Use logging when establishing or lowering row limit

Setting or lowering the row limit in an installation may cause a few existing SQL queries to show only the limited number of rows. Use logging to track usage that exceeds the limit. Investigate queries that exceed the default limit for correctness. If correct, use the limit parameter to prevent further warnings.

Additivity

Setting additivity to false prevents sql being logged to your main confluence log.

Example log

2012-10-20 05:14:57,019 [http-8116-4] userId: admin, dataSource: testDS, contentId: 91816371, spaceKey: zsql, title: dateRows, sql: DELETE from test_date;
            INSERT INTO test_date VALUES (1, date '2009-09-30', timestamp '2009-09-30 01:00', timestamp '2009-09-30 01:00', time '05:45:01', time '14:20');
            INSERT INTO test_date VALUES (2, date '2010-08-10', timestamp '2010-08-10 13:00', timestamp '2010-08-10 13:00', time '13:45:01', time '02:20');
            INSERT INTO test_date VALUES (3, date '2008-12-31', timestamp '2010-08-10 01:01', timestamp '2010-08-10 01:01', time '13:05:00', time '14:35');
            INSERT INTO test_date VALUES (4, date '2010-03-01', timestamp '2010-08-10 01:00', timestamp '2010-08-10 01:00', time '23:55:00', time '11:35');
            INSERT INTO test_date VALUES (5, date '2010-03-05', timestamp '2009-08-10 01:00', timestamp '2009-08-10 01:00', time '11:55:00', time '02:55');
            INSERT INTO test_date VALUES (6, date '2008-01-30', timestamp '2009-08-10 01:00', timestamp '2009-08-10 02:00', time '08:00:01', time '02:10');
            INSERT INTO test_date VALUES (7, date '2009-03-01', timestamp '2009-08-10 01:00', timestamp '2009-08-10 01:00', time '07:45:01', time '23:59');
            SELECT * from test_date;
2012-10-20 05:14:58,137 [http-8116-4] userId: admin, dataSource: testDS, contentId: 91816372, spaceKey: zsql, title: dateSort, sql: SELECT * from test_date

Timing logging

2012-10-21 19:29:48,113 [http-8116-4] statement time: 2293 ms, total time: 2300 ms, sql: select ...

How to log SQL activity

Log SQL statements

Add logging at runtime

Example log