JobOperator contains method to retrieve certain number of job data from the underlying job repository. When implement this in jdbc job repository, jberet currently retrieves all results from db, and get the subset as specified in JobOperator method. This can be slow or even blocking if a job repository contains large number of job data accumulated over time.

jdbc job repository should be able to get the required number of rows from db, by including the limit clause in the query. No all database products implement the portable way of specifying query limit, so ideally we should have slightly different sql statement for different db product. Currently jberet only includes a single shared jberet-sql.properties file.