<?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: Write in RAID5 disk space in Disk Enclosures</title>
    <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468366#M512</link>
    <description>Sorry, there is no way to write directly to RAID 5 on the AutoRAID. The AutoRAID actually writes first to the mirrored cache on the controlers and then copies the cache to RAID 10 on the disks. The only thing that would increase performance would be to add unallocated space to the AutoRAID allowing more RAID 10.</description>
    <pubDate>Wed, 29 Nov 2000 13:23:43 GMT</pubDate>
    <dc:creator>Bob Inglis</dc:creator>
    <dc:date>2000-11-29T13:23:43Z</dc:date>
    <item>
      <title>Write in RAID5 disk space</title>
      <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468365#M511</link>
      <description>I've a time problem when restoring to een autoraid. Restore take twice the time of a backup.&lt;BR /&gt;&lt;BR /&gt;Autoraid reserves a raid 0/1 space for write requests. If this space is full the autoraid copy data from raid 0/1 to raid 5. This action takes a lot of time. &lt;BR /&gt;Is it posible to write data direct into RAID5 space?&lt;BR /&gt;If possible does it give a better performance?&lt;BR /&gt;</description>
      <pubDate>Wed, 29 Nov 2000 10:25:57 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468365#M511</guid>
      <dc:creator>R. Berendsen</dc:creator>
      <dc:date>2000-11-29T10:25:57Z</dc:date>
    </item>
    <item>
      <title>Re: Write in RAID5 disk space</title>
      <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468366#M512</link>
      <description>Sorry, there is no way to write directly to RAID 5 on the AutoRAID. The AutoRAID actually writes first to the mirrored cache on the controlers and then copies the cache to RAID 10 on the disks. The only thing that would increase performance would be to add unallocated space to the AutoRAID allowing more RAID 10.</description>
      <pubDate>Wed, 29 Nov 2000 13:23:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468366#M512</guid>
      <dc:creator>Bob Inglis</dc:creator>
      <dc:date>2000-11-29T13:23:43Z</dc:date>
    </item>
    <item>
      <title>Re: Write in RAID5 disk space</title>
      <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468367#M513</link>
      <description>This may not be of any help to you, but I thought I would share my experience on this just in case.  Based on my personal experience, to really speed up the restore, the AutoRAID may need to be relatively empty at the initiation of the restore.  I am using OmniBack 3.0 for backup/restore.  I am on a K260 with 2 gig of memory, using a 2/15 DLT7000 library for the backups.&lt;BR /&gt;  My array had 12 18.2-drives in it.  I had 4 8-gig filesystems I blew away and restored as a test.  The filesystems were backed up using concurrency of 4 on the devices.  The restore took 8 hours.  Conversely, the same array (now with 8 18.2-gig drives and 4 36-gig drives and a total of 18 8-gigabyte mount points) was completely blown away, the volume groups recreated, and the data restored.  This ALSO took 8 hours.  I know that the AutoRAID writes directly to RAID5 when it detects a large sequential write, but I don't know what effect the elimination of the data (effectively creating a "well" to dump data into) on the array may have had.  I suspect that since the array wasn't empty the first time, it may have been busy moving stuff around while the restores were going on.</description>
      <pubDate>Wed, 29 Nov 2000 15:11:27 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468367#M513</guid>
      <dc:creator>Don Bentz</dc:creator>
      <dc:date>2000-11-29T15:11:27Z</dc:date>
    </item>
    <item>
      <title>Re: Write in RAID5 disk space</title>
      <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468368#M514</link>
      <description>Incidentally, I wouldn't want anybody to construe my comments as being an advisement to take the same approach to speeding up a restore.  I just thought it was a rather fascinating little study, but it's clearly a horrifying prospect to remove and restore all of your data.</description>
      <pubDate>Wed, 29 Nov 2000 15:20:27 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468368#M514</guid>
      <dc:creator>Don Bentz</dc:creator>
      <dc:date>2000-11-29T15:20:27Z</dc:date>
    </item>
    <item>
      <title>Re: Write in RAID5 disk space</title>
      <link>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468369#M515</link>
      <description>The AutoRAID itself desided to write direct to RAID5 if the file transfer limit is reashed.&lt;BR /&gt;For performance reason this make no different because of the Cache in the Diskarray from 96MB. The write performance is the same in the total amount.&lt;BR /&gt;The performance has no different because is backup software dependent. Normally the software is designed to make fast backup and slower restore, because we make more backup than restore.  &lt;BR /&gt;Best regards&lt;BR /&gt;Frank</description>
      <pubDate>Thu, 30 Nov 2000 08:35:32 GMT</pubDate>
      <guid>https://community.hpe.com/t5/disk-enclosures/write-in-raid5-disk-space/m-p/2468369#M515</guid>
      <dc:creator>Frank Skibbe</dc:creator>
      <dc:date>2000-11-30T08:35:32Z</dc:date>
    </item>
  </channel>
</rss>

