<?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 RAID performance with Oracle DB in Netservers</title>
    <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083381#M5491</link>
    <description>I'm using Oracle 8.1.6.3 on Windows2000Pro SP4.&lt;BR /&gt;I have a DB that's about 15GB. I have just restored the database to a fresh install on my LH 3 (2x400PII, 512MB, 8-10KRPM, 18.2GB drives in RAID5). I restored it from my laptop (P4-3.06GHz, 1GBRAM, 60GB, 5400 RPM). If I do a select count(*) on a 4.5 million record table on each machine, I get the following results.&lt;BR /&gt;Laptop - 10 Sec.&lt;BR /&gt;Netserver - ~45 Sec.&lt;BR /&gt;What could be causing the huge difference in DB response. I'm assuming it's a drive IO issue.</description>
    <pubDate>Wed, 01 Oct 2003 22:39:47 GMT</pubDate>
    <dc:creator>JimM_2</dc:creator>
    <dc:date>2003-10-01T22:39:47Z</dc:date>
    <item>
      <title>RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083381#M5491</link>
      <description>I'm using Oracle 8.1.6.3 on Windows2000Pro SP4.&lt;BR /&gt;I have a DB that's about 15GB. I have just restored the database to a fresh install on my LH 3 (2x400PII, 512MB, 8-10KRPM, 18.2GB drives in RAID5). I restored it from my laptop (P4-3.06GHz, 1GBRAM, 60GB, 5400 RPM). If I do a select count(*) on a 4.5 million record table on each machine, I get the following results.&lt;BR /&gt;Laptop - 10 Sec.&lt;BR /&gt;Netserver - ~45 Sec.&lt;BR /&gt;What could be causing the huge difference in DB response. I'm assuming it's a drive IO issue.</description>
      <pubDate>Wed, 01 Oct 2003 22:39:47 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083381#M5491</guid>
      <dc:creator>JimM_2</dc:creator>
      <dc:date>2003-10-01T22:39:47Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083382#M5492</link>
      <description>Your Laptop Configurations are HiFi to the Netserver you have in terms of Processor Speed and RAM.The processor speed and Ram are of very great configurations when compared to Netservers.So we can expect the difference in DB response.</description>
      <pubDate>Thu, 02 Oct 2003 00:10:38 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083382#M5492</guid>
      <dc:creator>Hari Kumar</dc:creator>
      <dc:date>2003-10-02T00:10:38Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083383#M5493</link>
      <description>The HDD on my laptop is a single 5400 rpm drive. The Array on the LH3 consists of eight 10K RPM, 18.2GB drives. The integrated NetRAID controller is set to Ultra-2. I can't expect to get better performance? Which would improve performance more? A pair of 600MHz PIIIs or a Gig of RAM?&lt;BR /&gt;The Processors in th LH3 only show about 25-30% untilization running the query.</description>
      <pubDate>Thu, 02 Oct 2003 07:35:38 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083383#M5493</guid>
      <dc:creator>JimM_2</dc:creator>
      <dc:date>2003-10-02T07:35:38Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083384#M5494</link>
      <description>Here One thing i want specify is RAID5 is the BEST in fault tolerance but PERFORMANE is a issue with it and we need plan carefully to span files accross different i/o and mirroring(control files, log files e.t.c ) to make I/O overheads limit.It gives better performance from cache and we must accept its performance at first read from disks.&lt;BR /&gt;A well approved writeup on keeping to RAID5&lt;BR /&gt;Please go thru this ----------&lt;BR /&gt;&lt;BR /&gt;RAID Performance Trade-offs&lt;BR /&gt;Conventional RAID demands a performance versus efficiency trade-off. This trade-off is&lt;BR /&gt;rooted in RAID Level 5, the most frequently implemented and commercially successful&lt;BR /&gt;variant of RAID. RAID 5 makes efficient use of disk space but often uses four distinct&lt;BR /&gt;write operations to put data onto disk. This is referred to as the â  RAID 5 Write Penalty.â  &lt;BR /&gt;This penalty consumes resources and impedes overall application performance.&lt;BR /&gt;A technique called the Block Pooling is used to avoid the RAID 5 Write Penalty. The&lt;BR /&gt;Block Pool is reserved space on disks that functions as â  virtual cacheâ   for disk write&lt;BR /&gt;processes. When data is to be written to RAID 5 disk, the RAID 5 Write Penalty is&lt;BR /&gt;avoided by first writing that data to the Block Pool using swifter RAID 0+1. Then in a&lt;BR /&gt;background process, relocates the data from the Block Pool to disk using the more space&lt;BR /&gt;efficient RAID 5. The Block Pool eliminates performance-robbing â  backpressureâ   on the&lt;BR /&gt;application from sluggish RAID 5 disk write operations. Figure 3 shows the interaction of&lt;BR /&gt;Concatenated Writes, Write Back Cache and the Bloc</description>
      <pubDate>Thu, 02 Oct 2003 08:53:37 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083384#M5494</guid>
      <dc:creator>Hari Kumar</dc:creator>
      <dc:date>2003-10-02T08:53:37Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083385#M5495</link>
      <description>Sorry the message could not be displayed properly.&lt;BR /&gt;Please find the attachment.&lt;BR /&gt;Hope that helps you.&lt;BR /&gt;Revert,&lt;BR /&gt;Thanks,</description>
      <pubDate>Thu, 02 Oct 2003 08:57:08 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083385#M5495</guid>
      <dc:creator>Hari Kumar</dc:creator>
      <dc:date>2003-10-02T08:57:08Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083386#M5496</link>
      <description>Hey,&lt;BR /&gt;Have been to the document ? Its provided  by HP !</description>
      <pubDate>Mon, 06 Oct 2003 10:52:25 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083386#M5496</guid>
      <dc:creator>Hari Kumar</dc:creator>
      <dc:date>2003-10-06T10:52:25Z</dc:date>
    </item>
    <item>
      <title>Re: RAID performance with Oracle DB</title>
      <link>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083387#M5497</link>
      <description>Jim,&lt;BR /&gt;&lt;BR /&gt;A pair of 600Mhz will give you better performance. You might also want to consider if you have ReadAhead/Normal settings in Netraid, cachedio,writeback/thru. what are you OS drives running on Software Raid/Hardware Raid, Is the Data on different Raid adapter/array.&lt;BR /&gt;Your laptop is definitely better in terms of processor speed. And it is running of only one single drive with no parity,unlike your server which has parity in disks plus it is flowing thru netraid adapter memory.&lt;BR /&gt;your laptop would have ddr memory which is faster than normal sdram memory.</description>
      <pubDate>Sat, 18 Oct 2003 00:09:09 GMT</pubDate>
      <guid>https://community.hpe.com/t5/netservers/raid-performance-with-oracle-db/m-p/3083387#M5497</guid>
      <dc:creator>Sachin_16</dc:creator>
      <dc:date>2003-10-18T00:09:09Z</dc:date>
    </item>
  </channel>
</rss>

