<?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 Re: XP -  performances. in Disk Enclosures</title>
    <link>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771347#M5563</link>
    <description>Hi,&lt;BR /&gt;&lt;BR /&gt;Instead of blockstriping try to use "extend-based" striping (with the default extentsize of 4 MB). &lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;Mark</description>
    <pubDate>Thu, 25 Jul 2002 05:44:02 GMT</pubDate>
    <dc:creator>Mark van Silfhout</dc:creator>
    <dc:date>2002-07-25T05:44:02Z</dc:date>
    <item>
      <title>XP -  performances.</title>
      <link>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771345#M5561</link>
      <description>xp512 with disc groups in raid 5.&lt;BR /&gt; 4 FC of high performance.&lt;BR /&gt; 8Gb memory cache.&lt;BR /&gt; lv's with striping of 4 discs and with block size of 256k.&lt;BR /&gt;dd with bs=256k gives 8Mb, the disc uses it to 12%. &lt;BR /&gt;5 dd bs=256k gives 80Mb/seg.&lt;BR /&gt;Oracle uses raw device with block size of 8k. &lt;BR /&gt;I need to improve much the performances.&lt;BR /&gt;  I need that oracle writes many GB in just a short time.&lt;BR /&gt; ideas ?</description>
      <pubDate>Wed, 24 Jul 2002 09:14:25 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771345#M5561</guid>
      <dc:creator>TOMAS BERNABEU</dc:creator>
      <dc:date>2002-07-24T09:14:25Z</dc:date>
    </item>
    <item>
      <title>Re: XP -  performances.</title>
      <link>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771346#M5562</link>
      <description>General rules:&lt;BR /&gt;&lt;BR /&gt;1. make your LV block size 64k.&lt;BR /&gt;2. use LDEVs (luns) from as many array groups as possible (stripe across AGs)&lt;BR /&gt;3. use the same # of luns from each ACP pair. (stripe across ACPs)&lt;BR /&gt;&lt;BR /&gt;Note that "dd" is not a good performance metric.&lt;BR /&gt;&lt;BR /&gt;Good luck</description>
      <pubDate>Wed, 24 Jul 2002 13:23:40 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771346#M5562</guid>
      <dc:creator>Vincent Fleming</dc:creator>
      <dc:date>2002-07-24T13:23:40Z</dc:date>
    </item>
    <item>
      <title>Re: XP -  performances.</title>
      <link>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771347#M5563</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;Instead of blockstriping try to use "extend-based" striping (with the default extentsize of 4 MB). &lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;Mark</description>
      <pubDate>Thu, 25 Jul 2002 05:44:02 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771347#M5563</guid>
      <dc:creator>Mark van Silfhout</dc:creator>
      <dc:date>2002-07-25T05:44:02Z</dc:date>
    </item>
    <item>
      <title>Re: XP -  performances.</title>
      <link>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771348#M5564</link>
      <description>- you have to look for best performance of server (cpu and memory and io !) and xp512 and oracle !&lt;BR /&gt;- if you have a xp512 i think you have an hp support engineer for xp-config&lt;BR /&gt;- use few disks (over all channels) only for oracle redo-logs, strip all other tablespaces over all other disks&lt;BR /&gt;- make "correct" application-tests for extend-based-stripping and ??K-stripping (16,32,64,..) and check the performance (not only io) with sar, glance, perfview,..&lt;BR /&gt;- check oracle-performance (statspack) and optimize init.ora</description>
      <pubDate>Fri, 26 Jul 2002 11:16:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/xp-performances/m-p/2771348#M5564</guid>
      <dc:creator>Wilfried Unkauf</dc:creator>
      <dc:date>2002-07-26T11:16:43Z</dc:date>
    </item>
  </channel>
</rss>

