<?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>topic EXAScaler Cloud by DDN: A shared file system to use with SAS Grid on Google Cloud Platform (GCP) in Administration and Deployment</title>
    <link>https://communities.sas.com/t5/Administration-and-Deployment/EXAScaler-Cloud-by-DDN-A-shared-file-system-to-use-with-SAS-Grid/m-p/743021#M22202</link>
    <description>&lt;P&gt;SAS recently tested SAS 9.4 on GCP instances using &lt;A href="https://www.ddn.com/cloud-services/" target="_blank"&gt;EXAScaler Cloud&lt;/A&gt; by DDN&lt;A href="https://console.cloud.google.com/marketplace/details/ddnstorage/exascaler-cloud?pli=1" target="_blank"&gt;: Google Marketplace - EXAScaler Cloud&lt;/A&gt;. The results of the testing were very favorable.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The test infrastructure consisted of 21 GCP instances defined like this:&amp;nbsp;&lt;/P&gt;
&lt;UL&gt;
&lt;LI&gt;Three n2-standard-32 instances for the SAS clients – each with 9 TB of local SSD storage.&lt;/LI&gt;
&lt;LI&gt;Sixteen n1-standard-16 instances for the OSS hosts – each with 4.3 TB of pd-balanced (balanced persistent disks) storage&lt;/LI&gt;
&lt;LI&gt;One n1-standard-4 instance for the MGS host – with 2 TB of pd-ssd (SSD persistent disks) storage&lt;/LI&gt;
&lt;LI&gt;One n1-standard-32 instance for the MDS host – with 1 TB of pd-balanced storage.&lt;/LI&gt;
&lt;LI&gt;EXAScaler 5.1.1 was used for the testing.&lt;/LI&gt;
&lt;/UL&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;For the test, the following EXAScaler tunings were applied:&lt;/P&gt;
&lt;OL&gt;
&lt;LI&gt;OSS tuning - Run on all OSS nodes&lt;/LI&gt;
&lt;/OL&gt;
&lt;P&gt;&lt;EM&gt;lctl set_param osd-ldiskfs.*.read_cache_enable=0 osd-ldiskfs.*.writethrough_cache_enable=0 obdfilter.*.brw_size=16 obdfilter.*.precreate_batch=1024&lt;/EM&gt;&lt;/P&gt;
&lt;OL start="2"&gt;
&lt;LI&gt;Lustre client tuning - Run on all client nodes&lt;/LI&gt;
&lt;/OL&gt;
&lt;P&gt;&lt;EM&gt;lctl set_param mdc.*.max_rpcs_in_flight=128 osc.*.max_pages_per_rpc=16M osc.*.max_rpcs_in_flight=16 osc.*.max_dirty_mb=1024 llite.*.max_read_ahead_mb=2048 osc.*.checksums=0 &amp;nbsp;llite.*.max_read_ahead_per_file_mb=256&lt;/EM&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;And the following file system tuning was applied:&lt;/P&gt;
&lt;P&gt;&lt;EM&gt;lfs setstripe -c -1 -S 4M &amp;nbsp;&amp;lt; file system subdirectory &amp;gt;&lt;/EM&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&lt;A href="https://support.sas.com/kb/59/680.html" target="_blank"&gt;&lt;STRONG&gt;rhel_iotest results&lt;/STRONG&gt;&lt;/A&gt;&lt;STRONG&gt;:&lt;/STRONG&gt;&lt;/P&gt;
&lt;P&gt;The 1-node test gave us 221&amp;nbsp;MB per second per physical core for WRITEs and 333 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The 2-node test gave us 226&amp;nbsp;MB per second per physical core for WRITEs and 337 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The 3-node test&amp;nbsp;gave us 230 MB per second per physical core for WRITEs and 325 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Please note that SAS recommends a minimum of 100 MB per second per physical core for WRITEs and READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;As you can see, the above EXAScaler Cloud infrastructure can produce the minimum IO throughput needs of a 64-core SAS Grid application.&amp;nbsp; Please reach out to DDN for more details on EXAScaler Cloud and how to configure it for optimal performance in GCP.&lt;/P&gt;</description>
    <pubDate>Fri, 21 May 2021 19:15:48 GMT</pubDate>
    <dc:creator>MargaretC</dc:creator>
    <dc:date>2021-05-21T19:15:48Z</dc:date>
    <item>
      <title>EXAScaler Cloud by DDN: A shared file system to use with SAS Grid on Google Cloud Platform (GCP)</title>
      <link>https://communities.sas.com/t5/Administration-and-Deployment/EXAScaler-Cloud-by-DDN-A-shared-file-system-to-use-with-SAS-Grid/m-p/743021#M22202</link>
      <description>&lt;P&gt;SAS recently tested SAS 9.4 on GCP instances using &lt;A href="https://www.ddn.com/cloud-services/" target="_blank"&gt;EXAScaler Cloud&lt;/A&gt; by DDN&lt;A href="https://console.cloud.google.com/marketplace/details/ddnstorage/exascaler-cloud?pli=1" target="_blank"&gt;: Google Marketplace - EXAScaler Cloud&lt;/A&gt;. The results of the testing were very favorable.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The test infrastructure consisted of 21 GCP instances defined like this:&amp;nbsp;&lt;/P&gt;
&lt;UL&gt;
&lt;LI&gt;Three n2-standard-32 instances for the SAS clients – each with 9 TB of local SSD storage.&lt;/LI&gt;
&lt;LI&gt;Sixteen n1-standard-16 instances for the OSS hosts – each with 4.3 TB of pd-balanced (balanced persistent disks) storage&lt;/LI&gt;
&lt;LI&gt;One n1-standard-4 instance for the MGS host – with 2 TB of pd-ssd (SSD persistent disks) storage&lt;/LI&gt;
&lt;LI&gt;One n1-standard-32 instance for the MDS host – with 1 TB of pd-balanced storage.&lt;/LI&gt;
&lt;LI&gt;EXAScaler 5.1.1 was used for the testing.&lt;/LI&gt;
&lt;/UL&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;For the test, the following EXAScaler tunings were applied:&lt;/P&gt;
&lt;OL&gt;
&lt;LI&gt;OSS tuning - Run on all OSS nodes&lt;/LI&gt;
&lt;/OL&gt;
&lt;P&gt;&lt;EM&gt;lctl set_param osd-ldiskfs.*.read_cache_enable=0 osd-ldiskfs.*.writethrough_cache_enable=0 obdfilter.*.brw_size=16 obdfilter.*.precreate_batch=1024&lt;/EM&gt;&lt;/P&gt;
&lt;OL start="2"&gt;
&lt;LI&gt;Lustre client tuning - Run on all client nodes&lt;/LI&gt;
&lt;/OL&gt;
&lt;P&gt;&lt;EM&gt;lctl set_param mdc.*.max_rpcs_in_flight=128 osc.*.max_pages_per_rpc=16M osc.*.max_rpcs_in_flight=16 osc.*.max_dirty_mb=1024 llite.*.max_read_ahead_mb=2048 osc.*.checksums=0 &amp;nbsp;llite.*.max_read_ahead_per_file_mb=256&lt;/EM&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;And the following file system tuning was applied:&lt;/P&gt;
&lt;P&gt;&lt;EM&gt;lfs setstripe -c -1 -S 4M &amp;nbsp;&amp;lt; file system subdirectory &amp;gt;&lt;/EM&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&lt;A href="https://support.sas.com/kb/59/680.html" target="_blank"&gt;&lt;STRONG&gt;rhel_iotest results&lt;/STRONG&gt;&lt;/A&gt;&lt;STRONG&gt;:&lt;/STRONG&gt;&lt;/P&gt;
&lt;P&gt;The 1-node test gave us 221&amp;nbsp;MB per second per physical core for WRITEs and 333 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The 2-node test gave us 226&amp;nbsp;MB per second per physical core for WRITEs and 337 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The 3-node test&amp;nbsp;gave us 230 MB per second per physical core for WRITEs and 325 MB per second per physical core for READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Please note that SAS recommends a minimum of 100 MB per second per physical core for WRITEs and READs.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;As you can see, the above EXAScaler Cloud infrastructure can produce the minimum IO throughput needs of a 64-core SAS Grid application.&amp;nbsp; Please reach out to DDN for more details on EXAScaler Cloud and how to configure it for optimal performance in GCP.&lt;/P&gt;</description>
      <pubDate>Fri, 21 May 2021 19:15:48 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Administration-and-Deployment/EXAScaler-Cloud-by-DDN-A-shared-file-system-to-use-with-SAS-Grid/m-p/743021#M22202</guid>
      <dc:creator>MargaretC</dc:creator>
      <dc:date>2021-05-21T19:15:48Z</dc:date>
    </item>
  </channel>
</rss>

