<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>question Re: YARN Tuning - How is memory overhead estimate determined? in Archives of Support Questions (Read Only)</title>
    <link>https://community.cloudera.com/t5/Archives-of-Support-Questions/YARN-Tuning-How-is-memory-overhead-estimate-determined/m-p/32004#M6771</link>
    <description>The value on the doc page is picked as about 20% of the RAM for overhead reservation, but you could set it lower. Our past overcommit testing does show that the values can reach close to extra 20% in use for some tested workloads, but that would not be an always-as-such case - and this may have changed overall lately also.&lt;BR /&gt;&lt;BR /&gt;We're reworking the docs for these recommendations soon in future, as developments happen. For now, please rely on the XLSX file for a more closer guideline on the recommended calculated values.</description>
    <pubDate>Fri, 18 Sep 2015 11:59:15 GMT</pubDate>
    <dc:creator>Harsh J</dc:creator>
    <dc:date>2015-09-18T11:59:15Z</dc:date>
    <item>
      <title>YARN Tuning - How is memory overhead estimate determined?</title>
      <link>https://community.cloudera.com/t5/Archives-of-Support-Questions/YARN-Tuning-How-is-memory-overhead-estimate-determined/m-p/30123#M6770</link>
      <description>&lt;P&gt;Looking at the &lt;A href="http://www.cloudera.com/content/cloudera/en/documentation/core/latest/topics/cdh_ig_yarn_tuning.html" target="_self"&gt;"Tuning YARN" documentation page&lt;/A&gt;, I noticed that the estimate for "Task Overhead" seems very large - 51GB for the example given, and 24GB on the tuning spreadsheet.&amp;nbsp; The s/s comment says: "Allow additional overhead for task buffers, such as the HDFS Sort I/O buffer, JVM overheads etc."&amp;nbsp; However, it does not explain how the number is derived.&amp;nbsp; This seems steep for a small/medium cluster that probably account for the majority of users.&amp;nbsp; Similar &lt;A href="http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.0.6.0/bk_installing_manually_book/content/rpm-chap1-11.html" target="_self"&gt;config guide at Hortonworks&lt;/A&gt; uses a rough ratio of 1/8 instead.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Please help - I need justification for requesting additional hardware $$s.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks,&lt;/P&gt;&lt;P&gt;Miles&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Fri, 16 Sep 2022 09:35:51 GMT</pubDate>
      <guid>https://community.cloudera.com/t5/Archives-of-Support-Questions/YARN-Tuning-How-is-memory-overhead-estimate-determined/m-p/30123#M6770</guid>
      <dc:creator>MilesYao</dc:creator>
      <dc:date>2022-09-16T09:35:51Z</dc:date>
    </item>
    <item>
      <title>Re: YARN Tuning - How is memory overhead estimate determined?</title>
      <link>https://community.cloudera.com/t5/Archives-of-Support-Questions/YARN-Tuning-How-is-memory-overhead-estimate-determined/m-p/32004#M6771</link>
      <description>The value on the doc page is picked as about 20% of the RAM for overhead reservation, but you could set it lower. Our past overcommit testing does show that the values can reach close to extra 20% in use for some tested workloads, but that would not be an always-as-such case - and this may have changed overall lately also.&lt;BR /&gt;&lt;BR /&gt;We're reworking the docs for these recommendations soon in future, as developments happen. For now, please rely on the XLSX file for a more closer guideline on the recommended calculated values.</description>
      <pubDate>Fri, 18 Sep 2015 11:59:15 GMT</pubDate>
      <guid>https://community.cloudera.com/t5/Archives-of-Support-Questions/YARN-Tuning-How-is-memory-overhead-estimate-determined/m-p/32004#M6771</guid>
      <dc:creator>Harsh J</dc:creator>
      <dc:date>2015-09-18T11:59:15Z</dc:date>
    </item>
  </channel>
</rss>

