<?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: MSA1000 and Partition alignment in MSA Storage</title>
    <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788953#M11285</link>
    <description>Watch out when using DISKPART to check the offsets, because it does not (always) show the correct value. I'll attach an example from Windows XP - don't have one from Windows server right now, but you can verify the results on your own box - feel free to report back.</description>
    <pubDate>Thu, 19 May 2011 10:07:43 GMT</pubDate>
    <dc:creator>Uwe Zessin</dc:creator>
    <dc:date>2011-05-19T10:07:43Z</dc:date>
    <item>
      <title>MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788948#M11280</link>
      <description>Hi all, &lt;BR /&gt; complicated question here, i need a very good disks expert... This is my challenge ;-)&lt;BR /&gt;&lt;BR /&gt;I'm trying to configure partition alignment on a MSA 1000 fw 7.20 Active/Active dual controller 512MB cache per controller.&lt;BR /&gt;&lt;BR /&gt;attached chassis fully populated.&lt;BR /&gt;2x msa30 shelf fully populated.&lt;BR /&gt;__________________________________&lt;BR /&gt;                 total: 42 Disks&lt;BR /&gt;&lt;BR /&gt;I'm following these specs:&lt;BR /&gt;&lt;A href="http://msdn.microsoft.com/en-us/library/dd758814%28v=sql.100%29.aspx" target="_blank"&gt;http://msdn.microsoft.com/en-us/library/dd758814%28v=sql.100%29.aspx&lt;/A&gt;&lt;BR /&gt;&lt;BR /&gt;and i noticed a thing on most of the LUNs:&lt;BR /&gt;on MSA1000 (ACU)&lt;BR /&gt;-stripe unit size: 128KB&lt;BR /&gt;-sectors: 32&lt;BR /&gt;&lt;BR /&gt;on windows&lt;BR /&gt;-cluster size: 128KB&lt;BR /&gt;-partition offset: 128KB&lt;BR /&gt;&lt;BR /&gt;dmdiag reveal a strange thing:&lt;BR /&gt;&lt;BR /&gt;---------- Partition Table Info Disk 2 ----------&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;           8,855 Cylinders&lt;BR /&gt;             255 Tracks/Cylinder&lt;BR /&gt;              63 Sectors/Track&lt;BR /&gt;             512 Bytes/Sector&lt;BR /&gt;              12 MediaType&lt;BR /&gt;     142,255,575 Sectors (total)&lt;BR /&gt;  72,834,854,400 Bytes (total)&lt;BR /&gt;      71,127,788 KB&lt;BR /&gt;          69,461 MB&lt;BR /&gt;            67.8 GB&lt;BR /&gt;&lt;BR /&gt;               0 StartingOffset&lt;BR /&gt;  72,837,857,280 PartitionLength&lt;BR /&gt;               0 HiddenSectors&lt;BR /&gt;               0 PartitionNumber&lt;BR /&gt;               0 PartitionType&lt;BR /&gt;               0 BootIndicator&lt;BR /&gt;               0 RecognizedPartition&lt;BR /&gt;               0 RewritePartition&lt;BR /&gt;&lt;BR /&gt;             MBR PartitionStyle&lt;BR /&gt;               4 PartitionCount&lt;BR /&gt;        21b45337 Signature&lt;BR /&gt;&lt;BR /&gt;        Starting        Partition     Hidden       Total  Partition    Partition       Boot Recognized    Rewrite&lt;BR /&gt;  Offset (bytes)   Length (bytes)    Sectors     Sectors     Number   Type (HEX)  Indicator  Partition  Partition&lt;BR /&gt;&lt;BR /&gt;         131,072   72,834,723,328        256 142,255,319          0         0x07          0          1          0&lt;BR /&gt;               0                0          0           0          1         0x00          0          0          0&lt;BR /&gt;               0                0          0           0          2         0x00          0          0          0&lt;BR /&gt;               0                0          0           0          3         0x00          0          0          0&lt;BR /&gt;&lt;BR /&gt;  72,834,854,400 Bytes (142255575 sectors) Geometric size&lt;BR /&gt;  72,837,857,280 Bytes (142261440 sectors) True size (measured)&lt;BR /&gt;  72,837,857,280 Bytes (142261440 sectors) Reported size (Partition0)&lt;BR /&gt;               0 Bytes (       0 sectors) missing/wasted&lt;BR /&gt;---------------------------------------------&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;dmdiag thinks sectors are 63, while MSA asserts they're 32... this could be the source of a misalignment!!! (am i wrong?)&lt;BR /&gt;&lt;BR /&gt;Wow, so i'm waiting for someone shading some light on the question... i'm stumped  -_-'&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;Thanks in advance to any brave sysAdmin trying to resolve this one... :-D&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 17 May 2011 14:28:47 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788948#M11280</guid>
      <dc:creator>Diego Castelli</dc:creator>
      <dc:date>2011-05-17T14:28:47Z</dc:date>
    </item>
    <item>
      <title>Re: MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788949#M11281</link>
      <description>I saw this also when I was playing with alignment.&lt;BR /&gt;&lt;BR /&gt;Even though I set LUN to 32, windows saw 63.&lt;BR /&gt;&lt;BR /&gt;In my notes, I say&lt;BR /&gt;  """&lt;BR /&gt;  ?????????????????????????????????&lt;BR /&gt;  SO WE MUST REBOOT to get 32s&lt;BR /&gt;  BUT it does NOT ALWAYS WORK ???????????????????&lt;BR /&gt;  May have to do a few reboots.&lt;BR /&gt;  ?????????????????????????????????&lt;BR /&gt;  """&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;But, it looks like your offset&lt;BR /&gt;&lt;BR /&gt;  131,072 = 64k&lt;BR /&gt;&lt;BR /&gt;is good.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;How did you partition this drive.&lt;BR /&gt;&lt;BR /&gt;When I did my alignment I used the GnuParted live CD and created a small, unused partition at the beginning of the drive with a size ~16MB such that it ends at aligned - 1.&lt;BR /&gt;The next partition starts at 16,777,216 = 256*64k&lt;BR /&gt;&lt;BR /&gt; (I was doing it mainly in a VMware Virtual Machine environment, so I had a lots of ways to play with it ;&amp;gt;)&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;bv&lt;BR /&gt;</description>
      <pubDate>Wed, 18 May 2011 17:52:33 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788949#M11281</guid>
      <dc:creator>Bob_Vance</dc:creator>
      <dc:date>2011-05-18T17:52:33Z</dc:date>
    </item>
    <item>
      <title>Re: MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788950#M11282</link>
      <description>Hi Bob, &lt;BR /&gt;i used Diskpart to partition, specifying the offset at 64K.&lt;BR /&gt;&lt;BR /&gt;I don't understand why do you think my offset is good even if sectors are 32?&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;32*512 bytes = 16384 bytes = 16KB&lt;BR /&gt;&lt;BR /&gt;63*512 bytes = 32256 bytes = 31,5 KB&lt;BR /&gt;&lt;BR /&gt;Even if MS asserts that moving 31,5 to 32 is good, i don't see how partition can be aligned if i have 32 sectors and 64KB offset. &lt;BR /&gt;Is that cause 64/16=4 (which is integer) ?&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Wed, 18 May 2011 22:33:32 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788950#M11282</guid>
      <dc:creator>Diego Castelli</dc:creator>
      <dc:date>2011-05-18T22:33:32Z</dc:date>
    </item>
    <item>
      <title>Re: MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788951#M11283</link>
      <description>&lt;!--!*#--&gt;As I read your output, the offset of the partition is&lt;BR /&gt;&lt;BR /&gt;  131072 = 128k&lt;BR /&gt;(my original post said =64k, which I meant to say 2*64k)&lt;BR /&gt;&lt;BR /&gt;and so partition is aligned on MSA stripe and the NTFS cluster size is a factor of the stripe size, so a cluser I/O never spans two stripes.&lt;BR /&gt;&lt;BR /&gt;It doesn't matter about the sectors as long as the offset is aligned.&lt;BR /&gt;The problem is getting the alignment done when Windows thinks there are 63 sector/trk.  It's insane today, when spt doesn't even mean anything, to be constrained by that math in Windows.&lt;BR /&gt;&lt;BR /&gt;I could not make diskpart work correctly when it thought spt=63, even though the MSA said 32.&lt;BR /&gt;That's why I asked how you did it.&lt;BR /&gt;So, that's why I went to GParted, which doesn't care.&lt;BR /&gt;Of course that required booting the CD, so Windows was down.&lt;BR /&gt;But, apparently, it worked for you.&lt;BR /&gt;&lt;BR /&gt;The reason that I also put the little, unused "alignment" partition at the beginning, is so it would show in Disk Mgt and would be a visual cue that the next partition is aligned.&lt;BR /&gt;&lt;BR /&gt;But, in reading the document that you linked in your first post, it's not clear about diskpart -- the document seemed to imply that it was a reporting issue.&lt;BR /&gt;Apparently, Windows is saying 63, but since it is *really* 32, diskpart is supposed to work?&lt;BR /&gt;&lt;BR /&gt;I use msinfo32.exe &lt;BR /&gt;        components/storage/disks&lt;BR /&gt;to dump partition info and that reported misalignment from a diskpart partition when spt=63, even when I specified a N*1024 offset.&lt;BR /&gt;spt=32 is OK because you can align it in normal N*1024 (64k, 128k, etc)&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;Bottom line:&lt;BR /&gt;It's still unclear to me, except that I know I can do it with GParted -- and Win2k8 always aligns correctly !&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;bv&lt;BR /&gt;</description>
      <pubDate>Thu, 19 May 2011 03:50:31 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788951#M11283</guid>
      <dc:creator>Bob_Vance</dc:creator>
      <dc:date>2011-05-19T03:50:31Z</dc:date>
    </item>
    <item>
      <title>Re: MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788952#M11284</link>
      <description>All right,&lt;BR /&gt;&lt;BR /&gt; I think the partition is always aligned with multiples of 16KB since 128K/(32*512B=16K)and  128K/(63*512B=31.5K, but is good to assume it to be 32K) are integer...&lt;BR /&gt;&lt;BR /&gt;Diskpart worked flawlessly (at least error-less) when i shooted the new partition command, specifying the &lt;BR /&gt;&lt;BR /&gt;Thanks for your help!&lt;BR /&gt;&lt;BR /&gt;I'm going to post another question, now for the speed since i noticed an Advisory and i'm now working in 160 MB/s but i don't know if it's normal....&lt;BR /&gt;&lt;BR /&gt;Hope you'll help there too, like here.&lt;BR /&gt;&lt;BR /&gt;Again many thanks.</description>
      <pubDate>Thu, 19 May 2011 09:29:01 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788952#M11284</guid>
      <dc:creator>Diego Castelli</dc:creator>
      <dc:date>2011-05-19T09:29:01Z</dc:date>
    </item>
    <item>
      <title>Re: MSA1000 and Partition alignment</title>
      <link>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788953#M11285</link>
      <description>Watch out when using DISKPART to check the offsets, because it does not (always) show the correct value. I'll attach an example from Windows XP - don't have one from Windows server right now, but you can verify the results on your own box - feel free to report back.</description>
      <pubDate>Thu, 19 May 2011 10:07:43 GMT</pubDate>
      <guid>https://community.hpe.com/t5/msa-storage/msa1000-and-partition-alignment/m-p/4788953#M11285</guid>
      <dc:creator>Uwe Zessin</dc:creator>
      <dc:date>2011-05-19T10:07:43Z</dc:date>
    </item>
  </channel>
</rss>

