<?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: Path Switching in Operating System - OpenVMS</title>
    <link>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007545#M79368</link>
    <description>No, the path switch cannot happen any quicker.&lt;BR /&gt;&lt;BR /&gt;We have not done any testing with EMC storage, so I cannot say for sure how those controllers respond, our experience with HSZ, HSG, and HSV controllers show us that&lt;BR /&gt;there can be some amount of time (measured in seconds) before the controller will return an error back to the operating system.&lt;BR /&gt;&lt;BR /&gt;The biggest consumer of time here is that we (multipath) need to wait for various SCSI commands to either "time out" or fail before we can move on to find a working path.  When an I/O fails, multipath's algorithms are such that we try to stay on the current path, because we've found that the majority of problems are rather transient.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;-- Rob, VMS Engineering</description>
    <pubDate>Mon, 09 Oct 2006 10:45:34 GMT</pubDate>
    <dc:creator>Robert Brooks_1</dc:creator>
    <dc:date>2006-10-09T10:45:34Z</dc:date>
    <item>
      <title>Path Switching</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007544#M79367</link>
      <description>I have been testing path swithes between Top and Bottom fibre cards on a ES45 running OpenVMS 7.3-2 with Update V8 applied.&lt;BR /&gt;We have put together a simple srcipt to test the time it takes to switch path. The script simply loops creating files on a EMC SAN based disk. While the script runs we simply pull a fibre cable from one of the fibre cards on the server.&lt;BR /&gt;The path switch takes approx 5 seconds.&lt;BR /&gt;Can this be speeded up ?&lt;BR /&gt;&lt;BR /&gt;Does a sysgen param exist to reduce the wait prior to path switch ?</description>
      <pubDate>Mon, 09 Oct 2006 10:14:17 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007544#M79367</guid>
      <dc:creator>Kevin Raven (UK)</dc:creator>
      <dc:date>2006-10-09T10:14:17Z</dc:date>
    </item>
    <item>
      <title>Re: Path Switching</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007545#M79368</link>
      <description>No, the path switch cannot happen any quicker.&lt;BR /&gt;&lt;BR /&gt;We have not done any testing with EMC storage, so I cannot say for sure how those controllers respond, our experience with HSZ, HSG, and HSV controllers show us that&lt;BR /&gt;there can be some amount of time (measured in seconds) before the controller will return an error back to the operating system.&lt;BR /&gt;&lt;BR /&gt;The biggest consumer of time here is that we (multipath) need to wait for various SCSI commands to either "time out" or fail before we can move on to find a working path.  When an I/O fails, multipath's algorithms are such that we try to stay on the current path, because we've found that the majority of problems are rather transient.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;-- Rob, VMS Engineering</description>
      <pubDate>Mon, 09 Oct 2006 10:45:34 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007545#M79368</guid>
      <dc:creator>Robert Brooks_1</dc:creator>
      <dc:date>2006-10-09T10:45:34Z</dc:date>
    </item>
    <item>
      <title>Re: Path Switching</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007546#M79369</link>
      <description>Thanks for the quick reply.&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 10 Oct 2006 04:42:17 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/path-switching/m-p/5007546#M79369</guid>
      <dc:creator>Kevin Raven (UK)</dc:creator>
      <dc:date>2006-10-10T04:42:17Z</dc:date>
    </item>
  </channel>
</rss>

