<?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: average I/O size ? in Operating System - OpenVMS</title>
    <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079729#M61413</link>
    <description>there has been tools which intercepted disk I/O and recorded information such as size of request. See for example EPM on Freeeware V5. I'm not aware of a standard tool.</description>
    <pubDate>Mon, 29 Sep 2003 04:05:47 GMT</pubDate>
    <dc:creator>Ian Miller.</dc:creator>
    <dc:date>2003-09-29T04:05:47Z</dc:date>
    <item>
      <title>average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079728#M61412</link>
      <description>We are currently running OpenVMS v7.3-1 using HSZ80 disk controllers.&lt;BR /&gt;&lt;BR /&gt;We can obtain much I/O info from the Monitor utility, but how does one determine the average disk I/O size?&lt;BR /&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Charlie</description>
      <pubDate>Fri, 26 Sep 2003 16:07:56 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079728#M61412</guid>
      <dc:creator>Charlie Powell</dc:creator>
      <dc:date>2003-09-26T16:07:56Z</dc:date>
    </item>
    <item>
      <title>Re: average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079729#M61413</link>
      <description>there has been tools which intercepted disk I/O and recorded information such as size of request. See for example EPM on Freeeware V5. I'm not aware of a standard tool.</description>
      <pubDate>Mon, 29 Sep 2003 04:05:47 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079729#M61413</guid>
      <dc:creator>Ian Miller.</dc:creator>
      <dc:date>2003-09-29T04:05:47Z</dc:date>
    </item>
    <item>
      <title>Re: average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079730#M61414</link>
      <description>&lt;BR /&gt;There really should be an MB/sec indicator somewhere in the standard tools no... but best I know there isn't. I would likely be maintained in the UCB (along with UCB$L_OPCNT), but I don't see it there. [ANAL/SYS... SHOW DEV DKxxx... READ IODEF ... FORMAT &lt;UCB&gt; ]&lt;BR /&gt;&lt;BR /&gt;Some performance tools (PSM, CA, PAWZ,... ?) will report this somehow.&lt;BR /&gt;&lt;BR /&gt;Anyway, In 25 years of VMS performance work I have had very little use for it. The bottleneck has always been IO/sec, not MB/sec because the average IO size has always been low 4KB.&lt;BR /&gt;&lt;BR /&gt;I believe theat the average IO size is drifting upwards 10KB? with RMS sequential &lt;BR /&gt;file buffers growing, RMS index file buckets being chosen larger, backups,... and I believe MB/sec can and will become an bottleneck.&lt;BR /&gt;&lt;BR /&gt;It looks like the new and expanded SHOW MEMORY/CACHE/FULL also puts the focus on IO/sec. Hmmm... I'll ask the engineer over lunch for details.&lt;BR /&gt;&lt;BR /&gt;You may be stuck using Hardware Monitor tools like VTDPY on the HSZ to get the KB/io numbers.&lt;BR /&gt;&lt;BR /&gt;Cheers,&lt;BR /&gt;Hein.&lt;BR /&gt;&lt;/UCB&gt;</description>
      <pubDate>Mon, 29 Sep 2003 09:55:28 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079730#M61414</guid>
      <dc:creator>Hein van den Heuvel</dc:creator>
      <dc:date>2003-09-29T09:55:28Z</dc:date>
    </item>
    <item>
      <title>Re: average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079731#M61415</link>
      <description>&lt;BR /&gt;Actually, while the base VMS does not keep track of IO sizes, the XFC does a great job in this space.&lt;BR /&gt;&lt;BR /&gt;Check out&lt;BR /&gt;HELP SHOW MEM /CACHE&lt;BR /&gt;SHOW MEM/CACHE/FULL&lt;BR /&gt;SHOW MEM/CACHE/VOL=xxx &lt;BR /&gt;&lt;BR /&gt;You'll even find nice histograms!&lt;BR /&gt;&lt;BR /&gt;Hein</description>
      <pubDate>Mon, 29 Sep 2003 15:09:39 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079731#M61415</guid>
      <dc:creator>Hein van den Heuvel</dc:creator>
      <dc:date>2003-09-29T15:09:39Z</dc:date>
    </item>
    <item>
      <title>Re: average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079732#M61416</link>
      <description>Try&lt;BR /&gt;ANALYZE /SYSTEM&lt;BR /&gt;(invokes the System Dump Analyzer SDA)&lt;BR /&gt;then see command SHOW.&lt;BR /&gt;Bruno</description>
      <pubDate>Tue, 30 Sep 2003 07:45:44 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079732#M61416</guid>
      <dc:creator>Bruno Ganino</dc:creator>
      <dc:date>2003-09-30T07:45:44Z</dc:date>
    </item>
    <item>
      <title>Re: average I/O size ?</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079733#M61417</link>
      <description>There is a utility in the HSZ80 called DSTAT which will give you (among other things) the average I/O transfer size. Simply type RUN DSTAT at the HSZ80&amp;gt; prompt during a busy period. I have attached Ken Bates' manual.&lt;BR /&gt;&lt;BR /&gt;There is a Windows analysis program (I think also written by Ken Bates) which can take the output of a long run of DSTAT and give you some meaningful stats. If your support staff won't send it to you, I can, but it's too big to post here.</description>
      <pubDate>Tue, 02 Dec 2003 00:16:14 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/average-i-o-size/m-p/3079733#M61417</guid>
      <dc:creator>Howard Taylor</dc:creator>
      <dc:date>2003-12-02T00:16:14Z</dc:date>
    </item>
  </channel>
</rss>

