<?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 VIOC, XFC and applications. in Operating System - OpenVMS</title>
    <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571368#M6306</link>
    <description>The discussion of MVDB made me think. For which databases, applications, ... do you better do set file/cache=no_cache ?&lt;BR /&gt;I'm thinking of Oracle, Sybase, DSM, SAS .... are any of these not bypassing the cache while they should because they cache data themselves ?&lt;BR /&gt;&lt;BR /&gt;Or did you experience big improvements when doing it for certain file activities ? &lt;BR /&gt;&lt;BR /&gt;Wim</description>
    <pubDate>Mon, 27 Jun 2005 08:11:43 GMT</pubDate>
    <dc:creator>Wim Van den Wyngaert</dc:creator>
    <dc:date>2005-06-27T08:11:43Z</dc:date>
    <item>
      <title>VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571368#M6306</link>
      <description>The discussion of MVDB made me think. For which databases, applications, ... do you better do set file/cache=no_cache ?&lt;BR /&gt;I'm thinking of Oracle, Sybase, DSM, SAS .... are any of these not bypassing the cache while they should because they cache data themselves ?&lt;BR /&gt;&lt;BR /&gt;Or did you experience big improvements when doing it for certain file activities ? &lt;BR /&gt;&lt;BR /&gt;Wim</description>
      <pubDate>Mon, 27 Jun 2005 08:11:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571368#M6306</guid>
      <dc:creator>Wim Van den Wyngaert</dc:creator>
      <dc:date>2005-06-27T08:11:43Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571369#M6307</link>
      <description>I'm told that setting DSM files nocache is a good thing.&lt;BR /&gt;See &lt;A href="http://dba.openvms.org/phorum/read.php?17,23" target="_blank"&gt;http://dba.openvms.org/phorum/read.php?17,23&lt;/A&gt;</description>
      <pubDate>Mon, 27 Jun 2005 10:39:52 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571369#M6307</guid>
      <dc:creator>Ian Miller.</dc:creator>
      <dc:date>2005-06-27T10:39:52Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571370#M6308</link>
      <description>I can confirm that DSM is not bypassing the cache in XFC.&lt;BR /&gt;&lt;BR /&gt;I also did some simple tests and found that Sybase 11 is not bypassing VIOC when doing simple operations  (create/drop table, insert).&lt;BR /&gt;&lt;BR /&gt;Could someone with XFC enabled and Oracle running check a disk with Oracle files on it with sh mem/cache=file=dev:*.*/out=xxx&lt;BR /&gt;and check xxx for presence of Oracle files (db, rollback segm, redo log) and report his findings (on my 7.3 the file selection of sh mem didn't work). Of course on a system where /cache=no has not been specified for the files.&lt;BR /&gt;&lt;BR /&gt;Or post the findings (esp. the hit rate of these files).&lt;BR /&gt;&lt;BR /&gt;Wim</description>
      <pubDate>Tue, 28 Jun 2005 06:09:41 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571370#M6308</guid>
      <dc:creator>Wim Van den Wyngaert</dc:creator>
      <dc:date>2005-06-28T06:09:41Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571371#M6309</link>
      <description>Wim: Some of the Oracle V8.17 files are cached. Even if some files are set no_cache a lot of Oracle files are cached. For instance on just one of the disks used by Oracle:&lt;BR /&gt;_DSA10:[ORACLE_V817.DBJAVA.LIB]CLASSES111.ZIP;3 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETCONFIG]BEQLSNR.EXE;2 _DSA10:[ORACLE_V817.DBJAVA]DBJAVAUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SVRMGR]SVRMGRL.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.INSTALL]FINDVMSV.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.UTIL]UTILUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SQLPLUS]GLOGIN.SQL;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.OCOMMON.NLS]LX10001.NLB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.DBJAVA]DETERMINE_INSTALLED_JDK.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SQLPLUS]CPYUS.MSB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK.ADMIN]TNSNAMES.ORA;40 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETCONFIG]TNSLSNR_64.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETCONFIG]TNSLSNR.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]SQLLDR.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_DATA.PCP]ORA_CONTROL2.CON;1 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETCONFIG]NETCONFIGUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ADABAS.DB013]PLOG.DAT;872 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.DBJAVA]JDBC_SETUP_JDK11.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK.LOG]LISTENER.LOG;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK.ADMIN]ORASRV_NETV2_PCP.COM;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]RDBMSCMD.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]OCIUS.MSB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK]NLUS.MSB;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]RDBMSUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.UTIL]ORAUSER.COM;3 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK]BEQLSNR0.COM;1 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.INSTALL]ORA_BEQ.DAT;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SQLPLUS]SQLPLUS.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.SVRMGR]SVRMGRUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.OCOMMON.NLS]LX00001.NLB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SQLPLUS]SQLPLUSUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_DATA.PCP]ORA_LOG1.RDO;1 (open)&lt;BR /&gt; Caching is enabled, active caching mode is _DSA10:[ORACLE_V817.NETWORK]ORA_FIND_BEQ_ID.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]ORACLIENT64_V817.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK]TNSLSNR.COM;1 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK]TNSUS.MSB;1 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.UTIL]ORACLIENT_V817.EXE;2 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETCONFIG]ANOUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.NETWORK.ADMIN]SQLNET.ORA;58 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.OCOMMON.NLS]LX20001.NLB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.INSTALL]INSTALLUSER.COM;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.OCOMMON.NLS]LX1BOOT.NLB;1 (closed)&lt;BR /&gt;_DSA10:[ORACLE_V817.SQLPLUS]SP2US.MSB;1 (closed)&lt;BR /&gt;_DSA10:[ADABAS.DB013]CLOG.DAT;8 (open)&lt;BR /&gt;_DSA10:[ORACLE_V817.RDBMS]ORACLE.EXE;2 (open)&lt;BR /&gt;Lawrence</description>
      <pubDate>Tue, 28 Jun 2005 15:19:41 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571371#M6309</guid>
      <dc:creator>Lawrence Czlapinski</dc:creator>
      <dc:date>2005-06-28T15:19:41Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571372#M6310</link>
      <description>All the files Lawrence shows should be cached. They do not contain data that Oracle woudl cache in its SGA. ThatÂ´s why Wim asked for dbf and redo files and such.&lt;BR /&gt;&lt;BR /&gt;fwiw,&lt;BR /&gt;Hei</description>
      <pubDate>Tue, 28 Jun 2005 19:39:52 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571372#M6310</guid>
      <dc:creator>Hein van den Heuvel</dc:creator>
      <dc:date>2005-06-28T19:39:52Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571373#M6311</link>
      <description>Lawrence,&lt;BR /&gt;&lt;BR /&gt;The file that are marked with no_cache are shown with "No Caching" on the 2nd line. So all files used are in the list.&lt;BR /&gt;&lt;BR /&gt;Note that pagefile.sys, swapfile and ssys$errorlog are marked as no cache by VMS itself.&lt;BR /&gt;&lt;BR /&gt;Could you check/post the redo log, rollback and database files ? Of course only files that have Total QIOs (in the output) at a high value.&lt;BR /&gt;&lt;BR /&gt;Thx&lt;BR /&gt;&lt;BR /&gt;Wim</description>
      <pubDate>Wed, 29 Jun 2005 01:25:06 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571373#M6311</guid>
      <dc:creator>Wim Van den Wyngaert</dc:creator>
      <dc:date>2005-06-29T01:25:06Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571374#M6312</link>
      <description>It might be a good idea on some systems to keep XFC active even if the db has a cache itself. E.g. on a system where there is a lot of free memory (e.g. for disaster reasons that everything must be able to run on 1 node of a 2 node cluster). Then a oracle cache of e.g. 100 MB would be extended by XFC to 600 MB.&lt;BR /&gt;&lt;BR /&gt;Only, when everything would be running on 1 node, the extended cache would be gone and performance could be (very) bad. So, ... it all depends.&lt;BR /&gt;&lt;BR /&gt;Wim</description>
      <pubDate>Wed, 29 Jun 2005 02:43:41 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571374#M6312</guid>
      <dc:creator>Wim Van den Wyngaert</dc:creator>
      <dc:date>2005-06-29T02:43:41Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571375#M6313</link>
      <description>Got information from Intersystems : DSM never bypasses XFC. SO, Ian was right.&lt;BR /&gt;&lt;BR /&gt;Wim</description>
      <pubDate>Wed, 29 Jun 2005 05:51:54 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571375#M6313</guid>
      <dc:creator>Wim Van den Wyngaert</dc:creator>
      <dc:date>2005-06-29T05:51:54Z</dc:date>
    </item>
    <item>
      <title>Re: VIOC, XFC and applications.</title>
      <link>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571376#M6314</link>
      <description>Interestly, mumps limits their cache size to 1 gb, so customers have told me they got better results with XFC on.&lt;BR /&gt;&lt;BR /&gt;The only way to know for sure is to test it.&lt;BR /&gt;&lt;BR /&gt;XFC just keeps getting better and better.  The major improvement is the speed it releases memory in tight situations.&lt;BR /&gt;&lt;BR /&gt;I usually recommend setting the db files /nocache but leaving XFC on.&lt;BR /&gt;&lt;BR /&gt;Warning XFC uses as much memory ss possible, so for best use, make sure you have a lot of memory.&lt;BR /&gt;&lt;BR /&gt;The see how many I/Os XFC is saving you type&lt;BR /&gt;show mem/cache/full &lt;BR /&gt;&lt;BR /&gt;and you will see the percentage of reads satisfied by cache.</description>
      <pubDate>Sat, 02 Jul 2005 04:18:38 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-openvms/vioc-xfc-and-applications/m-p/3571376#M6314</guid>
      <dc:creator>comarow</dc:creator>
      <dc:date>2005-07-02T04:18:38Z</dc:date>
    </item>
  </channel>
</rss>

