<?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 Modify SLO in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022855#M762550</link>
    <description>Hi, I've just configured WLM with vPars and I have a big doubt.&lt;BR /&gt;&lt;BR /&gt;I have the next scenario: &lt;BR /&gt;name           priority        CPUs&lt;BR /&gt;vpar0           1               2(now)/3(max)&lt;BR /&gt;vpar1           2               2(now)/3(max)&lt;BR /&gt;vpar2           3               2(now)/3(max)&lt;BR /&gt;&lt;BR /&gt;I have those 3 vpars at a 100% of usage with 2 cpus each one, so vpar0 takes 1 processor from vpar2 as it should.  &lt;BR /&gt;&lt;BR /&gt;Now the processor usage of vpar0 is near 230% (3 processors)and the SLO graphic shows that SLO is so near of the 50% (55% maybe). As I understand the graphic if the SLO is down that 50% (where the down limit is)it can return that processor, or if it goes over 75% where the top line is, it begins to search another processor (but by now the top is 3 processors) Is this supposition correct??&lt;BR /&gt;&lt;BR /&gt;vpar2 is working with only 1 processor (at 100%)and performance went down drastically. &lt;BR /&gt;&lt;BR /&gt;If my last supposition is correct how can I do the top limit of 75% goes to 80% and the down limit of 50% goes to 65%???&lt;BR /&gt;&lt;BR /&gt;Or how can I do that vpar0 returns that processor if it is not using more than 50% ??&lt;BR /&gt;&lt;BR /&gt;I used the vpar_usage_goal.wlm and vpar_usage_goal.wlmpar as a base for my complex.&lt;BR /&gt;&lt;BR /&gt;I hope anybody can help me.&lt;BR /&gt;&lt;BR /&gt;Regards.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;</description>
    <pubDate>Mon, 15 Jan 2007 14:39:39 GMT</pubDate>
    <dc:creator>David Islas González_1</dc:creator>
    <dc:date>2007-01-15T14:39:39Z</dc:date>
    <item>
      <title>Modify SLO</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022855#M762550</link>
      <description>Hi, I've just configured WLM with vPars and I have a big doubt.&lt;BR /&gt;&lt;BR /&gt;I have the next scenario: &lt;BR /&gt;name           priority        CPUs&lt;BR /&gt;vpar0           1               2(now)/3(max)&lt;BR /&gt;vpar1           2               2(now)/3(max)&lt;BR /&gt;vpar2           3               2(now)/3(max)&lt;BR /&gt;&lt;BR /&gt;I have those 3 vpars at a 100% of usage with 2 cpus each one, so vpar0 takes 1 processor from vpar2 as it should.  &lt;BR /&gt;&lt;BR /&gt;Now the processor usage of vpar0 is near 230% (3 processors)and the SLO graphic shows that SLO is so near of the 50% (55% maybe). As I understand the graphic if the SLO is down that 50% (where the down limit is)it can return that processor, or if it goes over 75% where the top line is, it begins to search another processor (but by now the top is 3 processors) Is this supposition correct??&lt;BR /&gt;&lt;BR /&gt;vpar2 is working with only 1 processor (at 100%)and performance went down drastically. &lt;BR /&gt;&lt;BR /&gt;If my last supposition is correct how can I do the top limit of 75% goes to 80% and the down limit of 50% goes to 65%???&lt;BR /&gt;&lt;BR /&gt;Or how can I do that vpar0 returns that processor if it is not using more than 50% ??&lt;BR /&gt;&lt;BR /&gt;I used the vpar_usage_goal.wlm and vpar_usage_goal.wlmpar as a base for my complex.&lt;BR /&gt;&lt;BR /&gt;I hope anybody can help me.&lt;BR /&gt;&lt;BR /&gt;Regards.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Mon, 15 Jan 2007 14:39:39 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022855#M762550</guid>
      <dc:creator>David Islas González_1</dc:creator>
      <dc:date>2007-01-15T14:39:39Z</dc:date>
    </item>
    <item>
      <title>Re: Modify SLO</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022856#M762551</link>
      <description>Not quite sure I understand what you are looking for but...&lt;BR /&gt;&lt;BR /&gt;There are min and max goal settings you can use.  i.e. don't take a CPU until I reach 90% and do not give back until I fall below some other value.  &lt;BR /&gt;&lt;BR /&gt;vPar priority also comes into play.  If all vPars are of equal priority then CPUs should go back to the origional systems once the added requirment falls below your thresholds.&lt;BR /&gt;&lt;BR /&gt;Look at the vPar config guide for CPU_min and max thresholds.&lt;BR /&gt;&lt;BR /&gt;This might lead you to what you are looking for.&lt;BR /&gt;</description>
      <pubDate>Mon, 15 Jan 2007 17:55:24 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022856#M762551</guid>
      <dc:creator>Tim Nelson</dc:creator>
      <dc:date>2007-01-15T17:55:24Z</dc:date>
    </item>
    <item>
      <title>Re: Modify SLO</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022857#M762552</link>
      <description>Hello.  &lt;BR /&gt;&lt;BR /&gt;Yes your understanding is correct.  By default, if the usage falls below 50%, it is assumed a CPU can be released.  If usage goes above 75%, more CPU is going to be requested.&lt;BR /&gt;&lt;BR /&gt;To change the usage goal thresholds from the default 50-75% for an SLO, add the desired bounds to the end of the 'usage _CPU' keyword.  For example in the vpar_usage_goal.wlm file:&lt;BR /&gt;&lt;BR /&gt;slo slo_myslo {&lt;BR /&gt;   pri = 1;  &lt;BR /&gt;   mincpu = 1;&lt;BR /&gt;   maxcpu = 700;&lt;BR /&gt;   goal = usage _CPU 65 80;  &amp;lt;&amp;lt;==&lt;BR /&gt;}&lt;BR /&gt;&lt;BR /&gt;One other thing to check is that cntl_base_previous_req is set to 1 (its default for WLM A.03.00 or later).  This tunable has to do with partitions releasing CPUs.  If you do not have it listed in your global tune structure, then it is safe to assume it is set to 1.</description>
      <pubDate>Tue, 16 Jan 2007 14:26:22 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022857#M762552</guid>
      <dc:creator>Jonathan Fears</dc:creator>
      <dc:date>2007-01-16T14:26:22Z</dc:date>
    </item>
    <item>
      <title>Re: Modify SLO</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022858#M762553</link>
      <description>Hi everybody, sorry about the late, I was out for some days.  I'm going to try this changes.&lt;BR /&gt;&lt;BR /&gt;Thanks a lot!!</description>
      <pubDate>Mon, 29 Jan 2007 14:02:04 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022858#M762553</guid>
      <dc:creator>David Islas González_1</dc:creator>
      <dc:date>2007-01-29T14:02:04Z</dc:date>
    </item>
    <item>
      <title>Re: Modify SLO</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022859#M762554</link>
      <description>Thanks a lot Jonathan it worked perfectly!!!&lt;BR /&gt;</description>
      <pubDate>Tue, 30 Jan 2007 17:36:55 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/modify-slo/m-p/5022859#M762554</guid>
      <dc:creator>David Islas González_1</dc:creator>
      <dc:date>2007-01-30T17:36:55Z</dc:date>
    </item>
  </channel>
</rss>

