![]() ![]() We also limit the service class in the where clause and this is because Amazon Redshift creates WLM query queues based on the service class. In analyzing a particular source, keeping the analysis to around a week’s time is usually enough to make some diagnosis of a problem. In the where clause, the date span is narrowed and that is to keep the load on the STL limited. service_class_start_time ) desc A Few Extra Notes service_class_start_time ) ORDER BY trunc ( w. total_exec_time ):: FLOAT ) AS "Percent WLM Queue Time" FROM stl_wlm_query w WHERE w. total_exec_time ) / 1000000 AS "Total Time", ( SUM ( w. total_exec_time ) / 1000000 AS "Total Exec Time", SUM ( w. total_queue_time ) / 1000000 AS "Total Queue Time", SUM ( w. service_class_start_time ) AS "Day", SUM ( w. Percent WLM Queue Time: This columns breaks down how long your queries were spending in the WLM Queue on the given day. ![]() Total Time: This column sums the previous two columns which will indicate how long it took for the queries on this source on the given day to return results to you. Total Exec Time: This column shows the total amount of time queries on the given day spent executing against the data source. Total Queue Time: This column shows the total amount of time queries on the given day spent waiting for an available connection on the source being analyzed. This query will have an output of five columns, and they are:ĭate: This column is the date on which the queries being analyzed were run. In this tutorial we will show you a fairly simple query that can be run against your cluster’s STL table analyzing the amount of time queries spend in your WLM’s queue and how long they execute against the source. Determining how much time your queries are spending either in the Workload Management (WLM) Queue or executing on your Amazon Redshift source can go a long way to improving your cluster’s performance. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |