<?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: OS read buffer size - I/O channel in Operating System - HP-UX</title>
    <link>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512808#M895988</link>
    <description>&lt;BR /&gt;If you have an 8k db block size in Oracle then you want to match this with the blocksize of the logical volume. &lt;BR /&gt;To find out what it is;&lt;BR /&gt;&lt;BR /&gt;fstyp -v  | grep bsize&lt;BR /&gt;&lt;BR /&gt;The normal blocksize for JFS filesystems is 8k anyway.&lt;BR /&gt;</description>
    <pubDate>Tue, 03 Apr 2001 10:30:26 GMT</pubDate>
    <dc:creator>Edward Sedgemore</dc:creator>
    <dc:date>2001-04-03T10:30:26Z</dc:date>
    <item>
      <title>OS read buffer size - I/O channel</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512807#M895987</link>
      <description>How do I determine the OS read buffer size on an HPUX system? I have an 8k db_block_size and want to correctly determine the db_multi_block_read_count parameter.</description>
      <pubDate>Tue, 03 Apr 2001 10:08:17 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512807#M895987</guid>
      <dc:creator>Raj Patel_1</dc:creator>
      <dc:date>2001-04-03T10:08:17Z</dc:date>
    </item>
    <item>
      <title>Re: OS read buffer size - I/O channel</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512808#M895988</link>
      <description>&lt;BR /&gt;If you have an 8k db block size in Oracle then you want to match this with the blocksize of the logical volume. &lt;BR /&gt;To find out what it is;&lt;BR /&gt;&lt;BR /&gt;fstyp -v  | grep bsize&lt;BR /&gt;&lt;BR /&gt;The normal blocksize for JFS filesystems is 8k anyway.&lt;BR /&gt;</description>
      <pubDate>Tue, 03 Apr 2001 10:30:26 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512808#M895988</guid>
      <dc:creator>Edward Sedgemore</dc:creator>
      <dc:date>2001-04-03T10:30:26Z</dc:date>
    </item>
    <item>
      <title>Re: OS read buffer size - I/O channel</title>
      <link>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512809#M895989</link>
      <description>Hi Raj,&lt;BR /&gt;&lt;BR /&gt;hfs block size is 8k and jfs is 1k , but may be this is not what you want :&lt;BR /&gt;&lt;BR /&gt;following is topic about db_multi_block_read_count &lt;BR /&gt;&lt;BR /&gt;---------------&lt;BR /&gt;It depends on what kind of application (DSS,OLTP) your database is running. Increasing DB_FILE_MULTIBLOCK_READ_COUNT doesnt always increase performance. Other factors like size of tables and whether parallel query is being utilized come into the picture. Unlike db_file_mult_read_cnt which is primarily for I/O performance, DB_BLOCK_SIZE affects other parameters like the maximum value of the 'freelists' storage param for tables and indexes. Increasing db_block_size also requires recreating your whole database with a larger block size. I dont think it's worthwhile to do that just for a small performance increase in I/O. So increasing one or the other is definitely not the same. I would increase the DB_file_mult... if I/O performace was an issue and test the results first. &lt;BR /&gt;--------------&lt;BR /&gt;this is from oracle metalink&lt;BR /&gt;&lt;BR /&gt;why not start at 16 ?&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 03 Apr 2001 10:31:01 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-hp-ux/os-read-buffer-size-i-o-channel/m-p/2512809#M895989</guid>
      <dc:creator>Printaporn_1</dc:creator>
      <dc:date>2001-04-03T10:31:01Z</dc:date>
    </item>
  </channel>
</rss>

