<?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: Swap space utilization in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071895#M735779</link>
    <description>&lt;!--!*#--&gt;Thanks guys.&lt;BR /&gt;&lt;BR /&gt;Here's an example from my test system:&lt;BR /&gt;&lt;BR /&gt;bigboy:(/tmp)(root)#UNIX95= ps -e -o vsz,uid,pid,ruser,args |sort -rn | head -7&lt;BR /&gt;   5364          0  2563 root     /usr/sbin/stm/uut/bin/tools/monitor/fpl_em&lt;BR /&gt;   4764         30  1726 www      /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;   4544          0  1519 root     /opt/samba/bin/smbd -D -s /etc/opt/samba/smb.conf&lt;BR /&gt;   4544          0  1489 root     /opt/samba/bin/smbd -D -s /etc/opt/samba/smb.conf&lt;BR /&gt;   4068          0  1208 root     /opt/dce/sbin/rpcd&lt;BR /&gt;   3740         30  1727 www      /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;   3484          0  1707 root     /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;&lt;BR /&gt;bigboy:(/tmp)(root)#swapinfo -mt&lt;BR /&gt;             Mb      Mb      Mb   PCT  START/      Mb&lt;BR /&gt;TYPE      AVAIL    USED    FREE  USED   LIMIT RESERVE  PRI  NAME&lt;BR /&gt;dev        4096       0    4096    0%       0       -    1  /dev/vg00/lvol2&lt;BR /&gt;reserve       -     141    -141&lt;BR /&gt;memory     1185     260     925   22%&lt;BR /&gt;total      5281     401    4880    8%       -       0    -&lt;BR /&gt;bigboy:(/tmp)(root)#&lt;BR /&gt;&lt;BR /&gt;If I get it right then although this system is not swapping anything, the summary of the numbers in the VSZ column from the 'ps' output(in kB) should be equal to the reserved MB value in the swapinfo output. Is that right?</description>
    <pubDate>Tue, 02 Oct 2007 08:07:52 GMT</pubDate>
    <dc:creator>Pavol Halcin</dc:creator>
    <dc:date>2007-10-02T08:07:52Z</dc:date>
    <item>
      <title>Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071892#M735776</link>
      <description>Hello, &lt;BR /&gt;&lt;BR /&gt;is there any way how to find out how much swap space a process has reserved/allocated?&lt;BR /&gt;I am facing a large swap space utilization on some systems and want to find out what application causes this behaviour. It's for sure being caused by SAP and Oracle, but I would like to see some exact numbers.&lt;BR /&gt;&lt;BR /&gt;Thanks for any helpful answer.</description>
      <pubDate>Tue, 02 Oct 2007 04:52:06 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071892#M735776</guid>
      <dc:creator>Pavol Halcin</dc:creator>
      <dc:date>2007-10-02T04:52:06Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071893#M735777</link>
      <description>&lt;BR /&gt;Hi ,&lt;BR /&gt;&lt;BR /&gt;Try this &lt;BR /&gt;&lt;BR /&gt;#UNIX95= ps -e -o vsz,uid,pid,ruser,args |sort -rn |more&lt;BR /&gt;&lt;BR /&gt;(one line command)&lt;BR /&gt;&lt;BR /&gt;Aneesh</description>
      <pubDate>Tue, 02 Oct 2007 05:17:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071893#M735777</guid>
      <dc:creator>Aneesh Mohan</dc:creator>
      <dc:date>2007-10-02T05:17:43Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071894#M735778</link>
      <description>Shalom,&lt;BR /&gt;&lt;BR /&gt;Overall:&lt;BR /&gt;swapinfo -tam&lt;BR /&gt;&lt;BR /&gt;&lt;A href="http://www.hpux.ws/?p=6" target="_blank"&gt;http://www.hpux.ws/?p=6&lt;/A&gt;&lt;BR /&gt;&lt;A href="http://www.hpux.ws/?p=8" target="_blank"&gt;http://www.hpux.ws/?p=8&lt;/A&gt;&lt;BR /&gt;&lt;BR /&gt;The second link includes a memory leak detector, which may help you spot a leak of ORACLE or SAP&lt;BR /&gt;&lt;BR /&gt;SEP</description>
      <pubDate>Tue, 02 Oct 2007 05:48:03 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071894#M735778</guid>
      <dc:creator>Steven E. Protter</dc:creator>
      <dc:date>2007-10-02T05:48:03Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071895#M735779</link>
      <description>&lt;!--!*#--&gt;Thanks guys.&lt;BR /&gt;&lt;BR /&gt;Here's an example from my test system:&lt;BR /&gt;&lt;BR /&gt;bigboy:(/tmp)(root)#UNIX95= ps -e -o vsz,uid,pid,ruser,args |sort -rn | head -7&lt;BR /&gt;   5364          0  2563 root     /usr/sbin/stm/uut/bin/tools/monitor/fpl_em&lt;BR /&gt;   4764         30  1726 www      /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;   4544          0  1519 root     /opt/samba/bin/smbd -D -s /etc/opt/samba/smb.conf&lt;BR /&gt;   4544          0  1489 root     /opt/samba/bin/smbd -D -s /etc/opt/samba/smb.conf&lt;BR /&gt;   4068          0  1208 root     /opt/dce/sbin/rpcd&lt;BR /&gt;   3740         30  1727 www      /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;   3484          0  1707 root     /opt/hpws/apache/bin/httpd -d /opt/hpws/apache -k start&lt;BR /&gt;&lt;BR /&gt;bigboy:(/tmp)(root)#swapinfo -mt&lt;BR /&gt;             Mb      Mb      Mb   PCT  START/      Mb&lt;BR /&gt;TYPE      AVAIL    USED    FREE  USED   LIMIT RESERVE  PRI  NAME&lt;BR /&gt;dev        4096       0    4096    0%       0       -    1  /dev/vg00/lvol2&lt;BR /&gt;reserve       -     141    -141&lt;BR /&gt;memory     1185     260     925   22%&lt;BR /&gt;total      5281     401    4880    8%       -       0    -&lt;BR /&gt;bigboy:(/tmp)(root)#&lt;BR /&gt;&lt;BR /&gt;If I get it right then although this system is not swapping anything, the summary of the numbers in the VSZ column from the 'ps' output(in kB) should be equal to the reserved MB value in the swapinfo output. Is that right?</description>
      <pubDate>Tue, 02 Oct 2007 08:07:52 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071895#M735779</guid>
      <dc:creator>Pavol Halcin</dc:creator>
      <dc:date>2007-10-02T08:07:52Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071896#M735780</link>
      <description>No, I'm afraid not.&lt;BR /&gt;&lt;BR /&gt;There are a few things which will complicate this sort of calculation:&lt;BR /&gt;&lt;BR /&gt;1) mmap() using MAP_FILE when the mapping is not MAP_PRIVATE (or is MAP_PRIVATE but all accesses are read). These objects will use virtual address space (and therefore show in VSZ [as they should]) but need no swap reservation because the object is backed by the file, not the swap area.&lt;BR /&gt;&lt;BR /&gt;2) Shared memory segments -- each segment will contribute to the VSZ of each attached process [i.e. a 1Mb (virtual) segment will contribute 1Mb of VSZ to each of processes A, B and C if each is attached].&lt;BR /&gt;&lt;BR /&gt;And probably some other nuances that are slipping my brain this early. What you really want is reported via pstat -- I think top and ps just don't have columns for it. Take a look at the attached program. Fair warnings/disclaimers: This is my own quick work, not production code. I haven't compiled it on anything older than r11.23 (though I think I handle prior releases... I just wouldn't want to bet the farm on it).&lt;BR /&gt;&lt;BR /&gt;If you compile 32-bit, I'd use the -D_PSTAT64 option... compile with +DD64 and you won't need to.&lt;BR /&gt;&lt;BR /&gt;The important field is the pst_swap field from the pst_vm_status structure for each object in the process(es) you care about... that's the swap reservation consumption. You should be able to use that to find your heavy hitters (and know if they're shared objects or whatnot).</description>
      <pubDate>Tue, 02 Oct 2007 08:26:23 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071896#M735780</guid>
      <dc:creator>Don Morris_1</dc:creator>
      <dc:date>2007-10-02T08:26:23Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071897#M735781</link>
      <description>Thanks a lot.&lt;BR /&gt;&lt;BR /&gt;I'll check your code on a test system and then consider the distribution to customer systems (won't be easy with over 1000 systems though).&lt;BR /&gt;&lt;BR /&gt;Regards</description>
      <pubDate>Tue, 02 Oct 2007 08:37:28 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071897#M735781</guid>
      <dc:creator>Pavol Halcin</dc:creator>
      <dc:date>2007-10-02T08:37:28Z</dc:date>
    </item>
    <item>
      <title>Re: Swap space utilization</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071898#M735782</link>
      <description>.</description>
      <pubDate>Tue, 02 Oct 2007 08:40:26 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/swap-space-utilization/m-p/5071898#M735782</guid>
      <dc:creator>Pavol Halcin</dc:creator>
      <dc:date>2007-10-02T08:40:26Z</dc:date>
    </item>
  </channel>
</rss>

