Okay.. they might be different; Just take a try; if the problem still there, please report a JIRA.
Francis Liang <[email protected]> 于2018年9月13日周四 下午4:48写道： > It says mapreduce.job.queuename config was fixed. Also other > configurations mentioned in the previous email were also fixed? Thanks! > Best, Feng. > > > > 发送自 Windows 10 版邮件 <https://go.microsoft.com/fwlink/?LinkId=550986>应用 > > > ------------------------------ > *发件人:* ShaoFeng Shi <[email protected]> > *发送时间:* Thursday, September 13, 2018 4:04:31 PM > *收件人:* user > *主题:* Re: mapreduce resource config does not work at ‘Convert Cuboid Data > to HFile’ Step > > Hi Feng, > > Please check KYLIN-3438. It has been fixed in 2.4.1 > > Francis Liang <[email protected]> 于2018年9月13日周四 下午3:59写道： > >> Hi: >> >> >> >> I am using kylin-2.4.0-bin-cdh57. For building a segment with more than >> 100m records, I need to overwrite some mapreduce configuration to bigger >> values. I did the following configuration at Cube level: >> >> >> >> kylin.engine.mr.config-override.mapreduce.map.java.opts=-Xmx10g >> >> kylin.engine.mr.config-override.mapreduce.map.memory.mb=10240 >> >> kylin.engine.mr.config-override.mapreduce.reduce.java.opts=-Xmx10g >> >> kylin.engine.mr.config-override.mapreduce.reduce.memory.mb=10240 >> >> >> >> When I was monitoring the resource usage in the cube building process, I >> found that the above configuration worked for every step except for the >> ‘Convert >> Cuboid Data to HFile’ Step. In this step, every container on Yarn was still >> using 1g memory and this caused the cube building task to fail. >> >> >> >> Could anyone help me on this? Thanks ahead! >> >> >> >> Best, Feng. >> >> >> >> 发送自 Windows 10 版邮件 <https://go.microsoft.com/fwlink/?LinkId=550986>应用 >> >> >> > > > -- > Best regards, > > Shaofeng Shi 史少锋 > > -- Best regards, Shaofeng Shi 史少锋
