Unique Constraint Errors in the Log Pertaining to RQL Due to Long Running Queries
Long running queries can fail to be recorded in a timely manner in the database.
If the error is similar to the form: org.hibernate.util.JDBCExceptionReporter - Violation of UNIQUE KEY constraint 'de_query_uq'. Cannot insert duplicate key in object 'amxbpm.de_query' then this is benign and can be ignored, providing the exception is reported by org.hibernate…, as the database has retried and succeeded in recording the query.
If the error is in addition reported by com.tibco.n2.de… then the error cannot be ignored. The retry mechanism has failed due to timeout.
If the retry mechanism fails then the following two properties may need to be adjusted in order to permit either more retry attempts or a longer period between retries.
ResourceQueryRetryLimit, default of 5
ResourceQueryRetryInterval, default of 500 milliseconds, or 0.5 seconds
By default the retry mechanism retries for a total of 2.5 seconds (ResourceQueryRetryLimit x ResourceQueryRetryInterval). This can be increased for example to 30 seconds by changing the values to:
ResourceQueryRetryLimit = 10
ResourceQueryRetryInterval = 3000