<?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: High Availabiligy for web application in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772036#M783170</link>
    <description>Just checking: have you made the weblogic plugins in the Web Servers to be cluster aware? (I.e. are they set up so they can connect to servers in either Geo as necessary?) Are the Java applications deployed _to the cluster_ instead of the individual Weblogic servers?&lt;BR /&gt;&lt;BR /&gt;If so, check the timeout values in weblogic plugin configuration. The default values may be  designed for non-cluster use: the plugin waits for a long time for any one WLS to answer, instead of rapidly timing out to try the other members of the cluster. &lt;BR /&gt;&lt;BR /&gt;At heavy loads, the long timeout will cause a huge backlog of requests waiting for a response, which may cause secondary problems (server running out of network sockets or some other resource).&lt;BR /&gt; &lt;BR /&gt;I think I saw this kind of phenomenon once, with Weblogic Server 7 or below, if I recall correctly.&lt;BR /&gt;&lt;BR /&gt;If the plugins are not configured as cluster-aware and you for some reason cannot make them so, you may need to switch over to the alternate Geo if Weblogic in the main Geo fails.&lt;BR /&gt;&lt;BR /&gt;How do you control which Geo is in use? If it's a load balancer of some sort, it will probably probe only the availability of the webserver tier, unless specifically configured to make test requests that get through the plugin to the Weblogic tier.&lt;BR /&gt;&lt;BR /&gt;</description>
    <pubDate>Mon, 17 Apr 2006 02:47:16 GMT</pubDate>
    <dc:creator>Matti_Kurkela</dc:creator>
    <dc:date>2006-04-17T02:47:16Z</dc:date>
    <item>
      <title>High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772032#M783166</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;We run our web application in 2 Geo. In case one Geo fails due to some reason another takes over.&lt;BR /&gt;However, if it doesn't happens in a proper order than it needs manual intervention. We call this&lt;BR /&gt;design as HA ( Highly Available ) solution. However, we only use Oracle 9i RAC on HPUX 11i under &lt;BR /&gt;Serviceguard as HA.&lt;BR /&gt;&lt;BR /&gt;I believe that it is more appropriate to use the word HA only when Operating System clustering&lt;BR /&gt;are used to run software products such as WebLogic or WebSphere or Web Servers etc.&lt;BR /&gt;&lt;BR /&gt;We use WebLogic Application cluster technology for maintaining redundancy, and availability  etc.&lt;BR /&gt;We observered failure of any instance of the weblogic clusters impacts other software &lt;BR /&gt;product like WebServer. For example failure or unavailability of WebLogic Managed Server instances causes Web&lt;BR /&gt;Servers to become hung because they are configured to talk through weblogic plugin.&lt;BR /&gt;&lt;BR /&gt;It seems like they maintain some sort of dependency.&lt;BR /&gt;&lt;BR /&gt;To overcome this situation i was considering to recommend operating system cluster such as &lt;BR /&gt;serviceguard. &lt;BR /&gt;&lt;BR /&gt;Any suggestion or idea from the gurus are appreciated.&lt;BR /&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Shiv</description>
      <pubDate>Fri, 14 Apr 2006 13:21:28 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772032#M783166</guid>
      <dc:creator>Shivkumar</dc:creator>
      <dc:date>2006-04-14T13:21:28Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772033#M783167</link>
      <description>Shalom Shiv,&lt;BR /&gt;&lt;BR /&gt;I would recommend a RAC-Serviceguard combination.&lt;BR /&gt;&lt;BR /&gt;If its already running on RAC and you add SG, you have the opportunity to double your capacity to handle traffic under normal circumstances.&lt;BR /&gt;&lt;BR /&gt;SG is the long time proven best OS clustering solution.&lt;BR /&gt;&lt;BR /&gt;It is not the only one. RedHat has a fairly robust Clustering suite of its own that I've go extensive experience with, if you wish to consisder that OS.&lt;BR /&gt;&lt;BR /&gt;I think moving away from weblogic is a good idea.&lt;BR /&gt;&lt;BR /&gt;SEP</description>
      <pubDate>Sat, 15 Apr 2006 17:35:06 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772033#M783167</guid>
      <dc:creator>Steven E. Protter</dc:creator>
      <dc:date>2006-04-15T17:35:06Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772034#M783168</link>
      <description>Shiv,&lt;BR /&gt;&lt;BR /&gt;I agree with SEP's idea,&lt;BR /&gt;&lt;BR /&gt;along with that I recon, if you are usng a Web servers, why not use then in N+1 base, If your requirement is 1, put 1 more in parallel. This will give more stability + added performance.&lt;BR /&gt;&lt;BR /&gt;Clustering is best for All, which N+1 is better for Web servers.&lt;BR /&gt;&lt;BR /&gt;Chan</description>
      <pubDate>Sun, 16 Apr 2006 08:03:23 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772034#M783168</guid>
      <dc:creator>Chan 007</dc:creator>
      <dc:date>2006-04-16T08:03:23Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772035#M783169</link>
      <description>SEP; &lt;BR /&gt;&lt;BR /&gt;We are using WebLogic as j2ee application server. Do you suggest moving it to Serviceguard ?&lt;BR /&gt;&lt;BR /&gt;We don't use Linux other than HPUX in our environment. HPUX is our favoured future direction.&lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;Shiv</description>
      <pubDate>Sun, 16 Apr 2006 16:12:28 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772035#M783169</guid>
      <dc:creator>Shivkumar</dc:creator>
      <dc:date>2006-04-16T16:12:28Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772036#M783170</link>
      <description>Just checking: have you made the weblogic plugins in the Web Servers to be cluster aware? (I.e. are they set up so they can connect to servers in either Geo as necessary?) Are the Java applications deployed _to the cluster_ instead of the individual Weblogic servers?&lt;BR /&gt;&lt;BR /&gt;If so, check the timeout values in weblogic plugin configuration. The default values may be  designed for non-cluster use: the plugin waits for a long time for any one WLS to answer, instead of rapidly timing out to try the other members of the cluster. &lt;BR /&gt;&lt;BR /&gt;At heavy loads, the long timeout will cause a huge backlog of requests waiting for a response, which may cause secondary problems (server running out of network sockets or some other resource).&lt;BR /&gt; &lt;BR /&gt;I think I saw this kind of phenomenon once, with Weblogic Server 7 or below, if I recall correctly.&lt;BR /&gt;&lt;BR /&gt;If the plugins are not configured as cluster-aware and you for some reason cannot make them so, you may need to switch over to the alternate Geo if Weblogic in the main Geo fails.&lt;BR /&gt;&lt;BR /&gt;How do you control which Geo is in use? If it's a load balancer of some sort, it will probably probe only the availability of the webserver tier, unless specifically configured to make test requests that get through the plugin to the Weblogic tier.&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Mon, 17 Apr 2006 02:47:16 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772036#M783170</guid>
      <dc:creator>Matti_Kurkela</dc:creator>
      <dc:date>2006-04-17T02:47:16Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772037#M783171</link>
      <description>Hi Matti, &lt;BR /&gt;&lt;BR /&gt;How to make weblogic plugin cluster aware ? I mean what config parameters do i need to configure into httpd.conf file of the apache ?&lt;BR /&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Shiv</description>
      <pubDate>Mon, 17 Apr 2006 18:01:16 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772037#M783171</guid>
      <dc:creator>Shivkumar</dc:creator>
      <dc:date>2006-04-17T18:01:16Z</dc:date>
    </item>
    <item>
      <title>Re: High Availabiligy for web application</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772038#M783172</link>
      <description>To make the plugin cluster-aware, you must not use WeblogicHost and WeblogicPort directives in the plugin configuration. Use the WeblogicCluster directive instead. For a two-node cluster with nodes named node1 and node2, listening in ports 7001 and 7002, the directive would be:&lt;BR /&gt;&lt;BR /&gt;WeblogicCluster node1:7001,node2:7002&lt;BR /&gt;&lt;BR /&gt;See also this BEA Weblogic documentation:&lt;BR /&gt;&lt;A href="http://e-docs.bea.com/wls/docs81/plugins/index.html" target="_blank"&gt;http://e-docs.bea.com/wls/docs81/plugins/index.html&lt;/A&gt;&lt;BR /&gt;&lt;BR /&gt;&lt;A href="http://e-docs.bea.com/wls/docs81/plugins/apache.html#113634" target="_blank"&gt;http://e-docs.bea.com/wls/docs81/plugins/apache.html#113634&lt;/A&gt;&lt;BR /&gt;(for Apache Web Server, if you use something else, see the equivalent chapter for your webserver)&lt;BR /&gt;&lt;BR /&gt;and particularly this diagram:&lt;BR /&gt;&lt;A href="http://e-docs.bea.com/wls/docs81/plugins/apache.html#132905" target="_blank"&gt;http://e-docs.bea.com/wls/docs81/plugins/apache.html#132905&lt;/A&gt;&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 18 Apr 2006 05:59:51 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/high-availabiligy-for-web-application/m-p/3772038#M783172</guid>
      <dc:creator>Matti_Kurkela</dc:creator>
      <dc:date>2006-04-18T05:59:51Z</dc:date>
    </item>
  </channel>
</rss>

