[ https://issues.apache.org/jira/browse/IOTDB-4561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17611326#comment-17611326 ]
Guanfei Guo commented on IOTDB-4561: ------------------------------------ In my PC, with single thread strategy, the computation costs about 10.8 seconds. In contrast, it only costs about 3.7 seconds with multi thread strategy. > Improve the RowRecord computing speed in GroupByWithoutValueFilterDataSet by > multi thread > ----------------------------------------------------------------------------------------- > > Key: IOTDB-4561 > URL: https://issues.apache.org/jira/browse/IOTDB-4561 > Project: Apache IoTDB > Issue Type: Improvement > Components: Core/Server > Reporter: Guanfei Guo > Assignee: Guanfei Guo > Priority: Critical > Labels: pull-request-available > Fix For: 0.13.3 > > > In rel/0.13, the computation of a RowRecord in > GroupByWithoutValueFilterDataSet is a single thread process. In case the > aggregation result cotaining hundreds of columns, the time for these > computation is much long (in my PC, about 10.8 seconds), so I use multi > thread strategy to speed up the computation. -- This message was sent by Atlassian Jira (v8.20.10#820010)