[ https://issues.apache.org/jira/browse/HADOOP-12356?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15064470#comment-15064470 ]
Wangda Tan commented on HADOOP-12356: ------------------------------------- [~cnauroth], bq. Am I right in thinking that NodeResourceMonitorImpl is the only caller of SysInfo#getCpuUsage (aside from test code)? If so, then SysInfo would be normalizing to [0, 100], only to have NodeResourceMonitorImpl immediately reverse the operation and change it back to core units. Would it be more helpful for SysInfo to return in core units in the first place? Agree, and we have getNumCores, and getNumProcessors. So we can normal cpu usage to 0-100% at anytime if we want. > CPU usage statistics on Windows > ------------------------------- > > Key: HADOOP-12356 > URL: https://issues.apache.org/jira/browse/HADOOP-12356 > Project: Hadoop Common > Issue Type: Bug > Components: util > Environment: CPU: Intel Xeon > OS: Windows server > Reporter: Yunqi Zhang > Assignee: Inigo Goiri > Labels: easyfix, newbie, patch > Attachments: 0001-Correct-the-CPU-usage-calcualtion.patch, > 0001-Correct-the-CPU-usage-calcualtion.patch, HADOOP-12356-v3.patch, > HADOOP-12356-v4.patch > > Original Estimate: 2h > Remaining Estimate: 2h > > The CPU usage information on Windows is computed incorrectly. The proposed > patch fixes the issue, and unifies the the interface with Linux. -- This message was sent by Atlassian JIRA (v6.3.4#6332)