<?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 Tru64 5.1 kernel memory management. in Operating System - Tru64 Unix</title>
    <link>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273633#M9792</link>
    <description>Hi,&lt;BR /&gt;That prior messages was a mistake...&lt;BR /&gt;&lt;BR /&gt;I'm having issues with the kernel memory allocator in Tru64 5.1 and IPFilter.&lt;BR /&gt;&lt;BR /&gt;On a 96MB AlphaSTATION 433, I can load IPFilter without any problems.  The exact same code compiled fails to load on a 256MB pc164 bsaed computer because of a 16k kernel memory bucket allocation failure.&lt;BR /&gt;&lt;BR /&gt;The relevant line of output from "vmstat -M" is:&lt;BR /&gt;bkt  elem_sz in_use  el_free bytes fail_nowait&lt;BR /&gt;  10   16384      3        4 49152       21&lt;BR /&gt;&lt;BR /&gt;I've not tried to tune any kernel parameters, this is just the default setup.&lt;BR /&gt;&lt;BR /&gt;Does anyone know what I need to do to fix this?&lt;BR /&gt;&lt;BR /&gt;</description>
    <pubDate>Tue, 11 May 2004 15:14:23 GMT</pubDate>
    <dc:creator>Darren Reed</dc:creator>
    <dc:date>2004-05-11T15:14:23Z</dc:date>
    <item>
      <title>Tru64 5.1 kernel memory management.</title>
      <link>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273633#M9792</link>
      <description>Hi,&lt;BR /&gt;That prior messages was a mistake...&lt;BR /&gt;&lt;BR /&gt;I'm having issues with the kernel memory allocator in Tru64 5.1 and IPFilter.&lt;BR /&gt;&lt;BR /&gt;On a 96MB AlphaSTATION 433, I can load IPFilter without any problems.  The exact same code compiled fails to load on a 256MB pc164 bsaed computer because of a 16k kernel memory bucket allocation failure.&lt;BR /&gt;&lt;BR /&gt;The relevant line of output from "vmstat -M" is:&lt;BR /&gt;bkt  elem_sz in_use  el_free bytes fail_nowait&lt;BR /&gt;  10   16384      3        4 49152       21&lt;BR /&gt;&lt;BR /&gt;I've not tried to tune any kernel parameters, this is just the default setup.&lt;BR /&gt;&lt;BR /&gt;Does anyone know what I need to do to fix this?&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 11 May 2004 15:14:23 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273633#M9792</guid>
      <dc:creator>Darren Reed</dc:creator>
      <dc:date>2004-05-11T15:14:23Z</dc:date>
    </item>
    <item>
      <title>Re: Tru64 5.1 kernel memory management.</title>
      <link>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273634#M9793</link>
      <description>please don't interpret message, post it in original context. Is your message "vmunix: malloc failed: bucket size"?&lt;BR /&gt;&lt;BR /&gt;If so the root cause of the problem could be: memory leak or wrong kernel parameters leading to an allocation problem.&lt;BR /&gt;&lt;BR /&gt;For a starter use sys_check and correct the parameters marked within the output.&lt;BR /&gt;</description>
      <pubDate>Tue, 11 May 2004 17:59:47 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273634#M9793</guid>
      <dc:creator>Ralf Puchner</dc:creator>
      <dc:date>2004-05-11T17:59:47Z</dc:date>
    </item>
    <item>
      <title>Re: Tru64 5.1 kernel memory management.</title>
      <link>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273635#M9794</link>
      <description>I'm with Ralf on this. Please start out with the exact (console?) message. Help us understand what made you worry about the 'kernel memory allocator'.&lt;BR /&gt;&lt;BR /&gt;Good call on the: vmstat -M,&lt;BR /&gt;but maybe you over-reduced the output.&lt;BR /&gt;If a next reply is needed, please attach the full output as a .txt file (maintains formatting).&lt;BR /&gt;&lt;BR /&gt;You can try increasing the map size (kmem_percent) and see if the problem goes away.  &lt;BR /&gt;sysconfig -q generic kmem_percent ?&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Wed, 12 May 2004 00:13:56 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273635#M9794</guid>
      <dc:creator>Hein van den Heuvel</dc:creator>
      <dc:date>2004-05-12T00:13:56Z</dc:date>
    </item>
    <item>
      <title>Re: Tru64 5.1 kernel memory management.</title>
      <link>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273636#M9795</link>
      <description>The output wasn't reduced (apart from vmstat -M): what's happening is MALLOC(9) is failing and there are no console messages from Mr vmunix.&lt;BR /&gt;&lt;BR /&gt;The actual code does something like this:&lt;BR /&gt;MALLOC(foo, void *, 16376, M_TEMP, M_NOWAIT)&lt;BR /&gt;&lt;BR /&gt;This should be ok, according to the man page for MALLOC(9).  Changing it to pass M_WAITOK for size &amp;gt; 4k has fixed the problem.&lt;BR /&gt;&lt;BR /&gt;But what I find strange is why a 96MB alpha works with M_NOWAIT and why the 256MB alpha didn't.  Is the kmem allocator different for the two variants?&lt;BR /&gt;&lt;BR /&gt;I suppose this is more a question for a forum related kernel (driver) development on Tru64 but I couldn't see one of those..&lt;BR /&gt;</description>
      <pubDate>Wed, 12 May 2004 10:35:36 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-tru64-unix/tru64-5-1-kernel-memory-management/m-p/3273636#M9795</guid>
      <dc:creator>Darren Reed</dc:creator>
      <dc:date>2004-05-12T10:35:36Z</dc:date>
    </item>
  </channel>
</rss>

