<?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: shmmax in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893243#M103418</link>
    <description>I don't believe shmmax set that high will hurt - if a program tries to grab an SHM segment when there isn't sufficient free memory, then that shmget call will just fail. The defunct processes are there because their parent(s) is not "wait"ing for them - when a child process dies, the parent must call wait to fetch the child's return status.  If the parent dies, then the children are adopted by init (process 1) which is repeatedly wait'ing for its children.  The only way, I see, that shmmax can in any way be related is if a parent forks a child that thinks it can grab a shm segment larger than your available memory (2 GB max), does not handle the shmget ret (fail) appropriately, dies or even exits gracefuly, and the parent does not wait for the dead child.</description>
    <pubDate>Thu, 30 Jan 2003 18:37:46 GMT</pubDate>
    <dc:creator>Byron Myers</dc:creator>
    <dc:date>2003-01-30T18:37:46Z</dc:date>
    <item>
      <title>shmmax</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893240#M103415</link>
      <description>Dear Admins..&lt;BR /&gt;&lt;BR /&gt;I have a K460 running HP-UX 11.00 64bit&lt;BR /&gt;Memory is 2GB&lt;BR /&gt;&lt;BR /&gt;the shmmax value is set : 3.75GB&lt;BR /&gt;shmmni is 512 and shmseg is 50.. is it ok to have shmmax so high?...&lt;BR /&gt;&lt;BR /&gt;The systems is running with UNIDATA.. it works fine except that i see a bunch of defunct processes. Is this related...&lt;BR /&gt;&lt;BR /&gt;please reccomend kernel changes id required..&lt;BR /&gt;Thanks in advance&lt;BR /&gt;George</description>
      <pubDate>Thu, 30 Jan 2003 18:27:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893240#M103415</guid>
      <dc:creator>George Abraham_1</dc:creator>
      <dc:date>2003-01-30T18:27:43Z</dc:date>
    </item>
    <item>
      <title>Re: shmmax</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893241#M103416</link>
      <description>Hi George,&lt;BR /&gt;&lt;BR /&gt;It's OK, but kind of pointless, to have it exceed the actual system RAM.&lt;BR /&gt;&lt;BR /&gt;I doubt the defunct processes are related to these values in any way.&lt;BR /&gt;&lt;BR /&gt;My $0.02,&lt;BR /&gt;Jeff</description>
      <pubDate>Thu, 30 Jan 2003 18:31:46 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893241#M103416</guid>
      <dc:creator>Jeff Schussele</dc:creator>
      <dc:date>2003-01-30T18:31:46Z</dc:date>
    </item>
    <item>
      <title>Re: shmmax</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893242#M103417</link>
      <description>George,&lt;BR /&gt;&lt;BR /&gt;I can't imagine any relationship between a high shmmax and defunct process - they're just process that didn't clean up after themselves when they died. A reboot or killing the parent (assuming it's not 1) will clear the defunct processes.&lt;BR /&gt;&lt;BR /&gt;You should be fine.&lt;BR /&gt;&lt;BR /&gt;Pete</description>
      <pubDate>Thu, 30 Jan 2003 18:34:26 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893242#M103417</guid>
      <dc:creator>Pete Randall</dc:creator>
      <dc:date>2003-01-30T18:34:26Z</dc:date>
    </item>
    <item>
      <title>Re: shmmax</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893243#M103418</link>
      <description>I don't believe shmmax set that high will hurt - if a program tries to grab an SHM segment when there isn't sufficient free memory, then that shmget call will just fail. The defunct processes are there because their parent(s) is not "wait"ing for them - when a child process dies, the parent must call wait to fetch the child's return status.  If the parent dies, then the children are adopted by init (process 1) which is repeatedly wait'ing for its children.  The only way, I see, that shmmax can in any way be related is if a parent forks a child that thinks it can grab a shm segment larger than your available memory (2 GB max), does not handle the shmget ret (fail) appropriately, dies or even exits gracefuly, and the parent does not wait for the dead child.</description>
      <pubDate>Thu, 30 Jan 2003 18:37:46 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893243#M103418</guid>
      <dc:creator>Byron Myers</dc:creator>
      <dc:date>2003-01-30T18:37:46Z</dc:date>
    </item>
    <item>
      <title>Re: shmmax</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893244#M103419</link>
      <description>your shmmax at 3.75G is the max for 11.00&lt;BR /&gt;&lt;BR /&gt;now if you want to go to 11.11, think much bigger&lt;BR /&gt;&lt;BR /&gt;depending on how much shared memory is requested, other programs may not have memory to start in&lt;BR /&gt;&lt;BR /&gt;a better number would probably be 1G</description>
      <pubDate>Thu, 30 Jan 2003 20:58:12 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/shmmax/m-p/2893244#M103419</guid>
      <dc:creator>John Bolene</dc:creator>
      <dc:date>2003-01-30T20:58:12Z</dc:date>
    </item>
  </channel>
</rss>

