<?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: looking for help:global sections that created by using sys$crmpsc lose data in Operating System - OpenVMS</title>
    <link>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864122#M34817</link>
    <description>likuair,&lt;BR /&gt;&lt;BR /&gt;  How are you closing down the program that writes the data? You can force data written by the current process to be written back to disk with the $UPDSEC service.&lt;BR /&gt;&lt;BR /&gt;  You can specify a flag to write all pages back. Also note the service is asynchronous.&lt;BR /&gt;&lt;BR /&gt; There's no simple means for making your program crash proof. Even if you update the section file after every write there, may still be timing windows where data can be lost. You may also find that the performance hit from having to update the disk after every update defeats the purpose of using a section file in the first place.&lt;BR /&gt;&lt;BR /&gt;Cheap, Fast, Reliable, pick TWO.</description>
    <pubDate>Sun, 17 Sep 2006 21:43:11 GMT</pubDate>
    <dc:creator>John Gillings</dc:creator>
    <dc:date>2006-09-17T21:43:11Z</dc:date>
    <item>
      <title>looking for help:global sections that created by using sys$crmpsc lose data</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864121#M34816</link>
      <description>i used sys$crmpsc to creat a section with a file named *.section,after install the section,i used sys$mgblsc to map data to the section.when i restart the alpha computer and reinstall the section,sometimes the data in the section was right,but sometimes the data was lost or wrong.&lt;BR /&gt;   what's the problem with my action?</description>
      <pubDate>Sun, 17 Sep 2006 21:18:19 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864121#M34816</guid>
      <dc:creator>likuair_1</dc:creator>
      <dc:date>2006-09-17T21:18:19Z</dc:date>
    </item>
    <item>
      <title>Re: looking for help:global sections that created by using sys$crmpsc lose data</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864122#M34817</link>
      <description>likuair,&lt;BR /&gt;&lt;BR /&gt;  How are you closing down the program that writes the data? You can force data written by the current process to be written back to disk with the $UPDSEC service.&lt;BR /&gt;&lt;BR /&gt;  You can specify a flag to write all pages back. Also note the service is asynchronous.&lt;BR /&gt;&lt;BR /&gt; There's no simple means for making your program crash proof. Even if you update the section file after every write there, may still be timing windows where data can be lost. You may also find that the performance hit from having to update the disk after every update defeats the purpose of using a section file in the first place.&lt;BR /&gt;&lt;BR /&gt;Cheap, Fast, Reliable, pick TWO.</description>
      <pubDate>Sun, 17 Sep 2006 21:43:11 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864122#M34817</guid>
      <dc:creator>John Gillings</dc:creator>
      <dc:date>2006-09-17T21:43:11Z</dc:date>
    </item>
    <item>
      <title>Re: looking for help:global sections that created by using sys$crmpsc lose data</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864123#M34818</link>
      <description>I have also tried using sys$updsecw after mapping data,but the problem has still happened.&lt;BR /&gt;   And the flag that I used was &lt;BR /&gt;   flags = SEC$M_WRT | SEC$M_GBL;&lt;BR /&gt;   Are there really no means to sovle the problem?&lt;BR /&gt;   maybe I can creat a new file named b.dat,and every hour write data that storing in the *.section to b.dat.When the problem happened ,I can rewrite the data in b.dat back to *.section.Is it OK?</description>
      <pubDate>Sun, 17 Sep 2006 22:22:23 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864123#M34818</guid>
      <dc:creator>likuair_1</dc:creator>
      <dc:date>2006-09-17T22:22:23Z</dc:date>
    </item>
    <item>
      <title>Re: looking for help:global sections that created by using sys$crmpsc lose data</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864124#M34819</link>
      <description>&lt;BR /&gt;Our database makes extensive use of global sections.  When the application has a clean shutdown there's no issues, but when a system is restarted after a crash, we have to rebuild any global sections.  This leads to additional downtime after a crash.  &lt;BR /&gt;&lt;BR /&gt;We write files to signal normal shutdown/restart or unknown state rebuild required to restart proceedures.  &lt;BR /&gt;&lt;BR /&gt;As John points out, writing to disk negates many of the speed advantages using global sections.  &lt;BR /&gt;&lt;BR /&gt;Andy</description>
      <pubDate>Mon, 18 Sep 2006 01:55:18 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/looking-for-help-global-sections-that-created-by-using-sys/m-p/3864124#M34819</guid>
      <dc:creator>Andy Bustamante</dc:creator>
      <dc:date>2006-09-18T01:55:18Z</dc:date>
    </item>
  </channel>
</rss>

