Storage – Strip Unit Size


A Storage disk’s Strip Unit Size is often taken into consideration as one considers storage alignment.

So I am finishing up and closing all opened Google’s Chrome Tab, and went back and read:

Disk Partition Alignment (Sector Alignment) for SQL Server: Part 4: Essentials (Cheat Sheet)

Jimmy May’s paper principally deals with Microsoft products: SQL Server on Windows OS, but his formula is very generalized.


Strip Unit Size – What does it matter?

Let us see how it plays out:

Here is Jimmy May’s formula:

Three Values - Two Essential Correlations

Jimmy’s document says:

  • Perform these calculations for each partition which must result in integer values
  • Of the two, the first is far more important.  Use the information below to divine this information.

And, here is how he says to get “File Allocation Unit Size” and also “Starting Offset”:


It is left to the inquiring mind how to get “Stripe Unit Size” as that number is not based on the OS, Microsoft Windows, in this case; but based on the Vendor and each Application.

Matrix – Vendor

Here are some numbers for various vendors and models:

Strip Unit Size
Vendor Model# Strip Unit Size
EMC Clarion  64k
NetApp (all models)  4k
HP (all models)  64k -to- 256k

Matrix – Application

Here are popular applications along with their Block Size.

Block Size
Vendor Application Block Size
Microsoft SQL Server  64 kb
Hadoop HBASE  64 kb
MySQL\Percona InnoDB  512 Bytes

Matrix – Application – MySQL/Inno DB

This variable changes the size of transaction log records. The default size of 512 bytes is good in most situations. However, setting it to 4096 may be a good optimization with SSD cards. While settings other than 512 and 4096 are possible, as a practical matter these are really the only two that it makes sense to use. Clean restart and removal of the old logs is needed for the variable innodb_log_block_size to be changed.


Calculation – NetApp

  1. NetApp’s default Strip Unit Size is 4K and Microsoft’s SQL Server best practice suggest using a File Allocation Unit Size of 64K.
  2. It does not take much calculation to deduce that 4K / 64K will not render an integer value, but a fractional value
  3. Please keep in mind that for NTFS, the default Strip Unit Size is 4K and one will get a whole number of 1 (Strip Unit Size / File Allocation Unit Size = 4 K / 4K = 1)


References – Vendor – Microsoft

References – Vendor – NetApp

References – Vendor – EMC

References – Vendor – HP

References – Vendor – Oracle

References – Vendor – MySQL

References – Technology – Hadoop

References – Changing Block Size

Microsoft – SQL Server – Storage – Disk Alignment (via DiskPart Scripting)


As part of a Storage I/O review, discussions whether our disks are properly aligned came back up.

The specific Knowledge Base article that I was pointed to is :

How to diagnose misaligned I/O on Windows hosts

It is a recently published NetApp article; as its publish date is 2013.02.27.

Though a short article, there is a lot in it.

There are two areas that I will like to cover in this posting; those areas are StartingOffset and Partition Style.

Starting Offset:

Depending on the Version of Windows (OS Version) , Partition Style, and your LUN size, your gold standard for “Starting offset” appears to vary a bit.

Here are the Numbers published by NetApp:

  • For Windows MBR, this number should be 32256, 31.5kb offset is used when the LUN is created with the Windows LUN type (31.5 * 1024 = 32256).
  • For Windows GPT, this number should be 65535 bytes for LUNs smaller than 4GB or 1048576 bytes for LUNs that are 4GB or larger.
  • For Windows 2008+, this number should be 65535 bytes for LUNs smaller than 4GB or 1048576 bytes for LUNs that are 4GB or larger

In summary:

  • Windows MBR, 32556
  • Windows GPT, disk size < 4 GB, then 65535 : else disk size >=  4GB, then 1048576


There are a couple of procedures you can employ to determine your Starting Offset:

  • Windows Management Interface (WMI)



wmic partition get BlockSize, StartingOffset, Name



OS version

Starting Offset

  • On the disk we are most interested in, Disk 4, our Starting Offset is 1048576.  That number matches up with NetApp guidance.

Partition Style:

Microsoft supports a couple of partition styles, MBR and GPT.

  • MBR — Legacy partition
  • GPT – new partition

There a couple of ways to determine your partition style:

  • GUI – Disk Management
  • DiskPart / list disk

Partition Style – Detect via Disk Management

Here are the steps:

  • Launch “Computer Management”
  • On the left panel, transverse to “Storage” \ “Disk Management”
  • The list of available disks are displayed on the right panel
  • On the right panel, select the disk you are interested in — make sure you select the physical disk, and not logical disk
  • Right click on your selected physical disk, and select “Properties” from the drop-down  menu
  • Notice that the name of the window that shows up will very based on whether this is a local disk, a SAN disk and the Disk Vendor
  • Proceed to the “Volumes” tab

Here is what shows up for us:

NetApp Device Properties

Partition Style – Via DiskPart / listdisk

Here are the steps:

  • Launch OS Shell
  • Start diskpart interactively (diskpart)
  • Issue “List Disk

DiskPart - List disk


To determine which disks are GPT, follow the GPT column.

If a disk has the asterisk symbol, then it is GPT. Else, it is not…

The two checks we performed conclusively affirm that our “Disk 4” is in fact MBR.  How could this be:

So went back and looked at our scripts:

select disk 4
create partition primary align=1024
assign letter=V
format fs=ntfs unit=64k label="Disk - Temp" quick

The script looks good:


Googled some more and found out what is wrong:

  • The Create partition syntax does not allow us to directly set the Partition Style


To correct your script for the future, have it resemble something along the likes of:

select disk 4
convert gpt
offline disk
online disk
attribute disk clear readonly
create partition primary align = 1024
assign letter=V
format fs=ntfs unit=64k label="Disk - Temp" quick

Please show and demonstrate extraordinary care when preparing to issue the script above:

  • Make sure you have selected the right disk (Select disk 4)
  • Notice the use of “clean” — It destroys the disk
  • Convert gpt –> Converts the partition to gpt; the default is mbr
  • Notice the use of “attribute disk clear readonly”; It says to the disk remove the armor you place on formatted and in-use disk

Why GPT:


Is there a performance penalty with choosing either MBR or GPT as your Partition Style. Googled for help, and there does not appear to be.

Please keep in mind that the partition style does not affect the way your data is written out, it has more to do with your partition table.

And, Microsoft will not let you get away with a wrong choice (ie MBR) for disks bigger than 2TB.




References – GPT ( Processing)

References – Vendor

References – Vendor – NetApp:

Technical: NetApp – MPIO – Path Details

Technical: NetApp – MPIO – Path Details

As part of NetApp diagnostic, you might need to dig deep into which paths are actually being used.

MPIO Path Details

  1. Launch Computer Management
  2. In the left panel, Access Storage \ Data OnTap(R) DSM Management \ Disk Management
  3. In the right panel, select the Disk
  4. Right Click on the the Disk Nth, and in the ensuring “Drop-down” menu, select the “Properties”
  5. The “NETAPP LUN Multi-Path Disk Device Properties” window appears
  6. The paths are listed in the “This device has the following paths”
  7. Double-clicked on the path you want to dig into …
  8. The “MPIO Path Details” window appears

NetApp - MPIO Path Details

The following areas are displayed:

  • Number of Reads
  • Number of Writes
  • Bytes Read
  • Bytes Written


Note that in the instructions above, you can not select the “Logical Disk”.


Computer Management


So in the screen above, please select “Disk 4” and right click on that selection.


NetApp – Data Collection Tool for Windows (ONTAPWinDC.exe)

Microsoft Visual C++ 2008

Ran OnTapWinDC, but unfortunately received an error message:

The application has failed to start because its side-by-sideconfiguration  is in correct. Please see the application 
event log or use the command-line sxstrace.exe tool for more detail.

To fix download and install “Microsoft Visual C++ 2008 SP1 redistributable package”. For X64 versions, download the following:

I installed “Microsoft Visual C++ 2008 SP1”.  And, re-ran OnTapWinDC.

And, my God it launched successfully.

But, when I tried gathering data it shows the error listed below: Authentication failed <Filer Name>

NetApp - OnTapWinDC

I spent all evening trying to fix this one. Missed the last bus and everything in between.

I knew I had a problem with Network Firewall and not just simple user-name & password authentication.

Tried setting up a SSH Tunneling with MS Windows based Free SSH Servers.  Had one working with Cygwin, but was not in a mood to set up another one.

Nevertheless, finally, resorted to using a pre-existing Linux one.

To create the SSH Tunnel that will facilitate SSH Access to the Filers:



  plink [SSHServer] -P 22 -C -L [BridgeAddress]:BridgePort:DestinationAddress:22


 plink SSHServer -ssh -P 22 -C -L

To ensure that your connections are set up, use putty and see if you ‘re able to connect to the Filers:

  • Run putty
  • Target hostname : localhost
  • Target Port number: 22

Putty screen:

putty - localhost - port 22

The system came back with Filer asking me to authenticate.  Should have taken the help of the best NetApp engineer in world and configured Filer’s SSH Key authentication.

And, would not have to enter passwords.  Issued the password and we are good.

Issue a couple of NetApp Host Commands:



netapp - validation

Run netstat -ano to review your Network Connections:

netstat -ano | find ":22 "


Netstat -ano | find ":22 "

netstat - port - 22 - port 22

And, then returned to the tool:

  • Filer Name/IP –> Make sure to use localhost  — as ssh tunneling will redirect

NetApp - OnTapWinDC - User Authentication - post ssh

But, still no luck.  The same error message:Authentication Failed.

I am here thanking Craig:

Gotcha with NetApp’s OnTapWinDcTool

Basically, he says that connectivity to Filer is actually over HTTP, and not SSH. And, that to fix you should connect to your Filer and enable HTTP.

options httpd.admin.enable on

To facilitate HTTP Access over TCP/IP Tunneling:

  plink SSHServer -ssh -P 22 -C -L

To validate our SSH Connection, issue:

   netstat -ano | find ":80"

And, you should see at minimum the following entries:

netstat - port - 80

The data is easy enough to read:

  • Column – 1 – Protocol {TCP}
  • Column – 2 – Local IP Address and Port Number { LocalHost:80)
  • Column – 3 – Destination IP Address and Port Number
  • Column – 4 – Status { Listening}
  • Column – 5 – Process ID {13080}

The Process ID is important.  To terminate the SSH Connection, issue a kill request directed to the process ID.

Once we have SSH Tunneling to the Filer over port 80, we were good.


References – NetApp – OnTapWinDCTool

References – SSH Tunneling

NetApp – Performance and Statistics Collector (PerfStats)

NetApp – Performance and Statistics Collector (PerfStats) – version 7

Command Line Parameters


[-f controllername[,controllername1,controllername2,...]]
  • name of the filer



 [-t time] (sample time per iteration, default 2)
  • Duration of each iteration in minutes


[-i n[,m]] (repeat n times with m minutes between samples, 
                     defaults: n=1,m=0)
  • Number of Iterations and wait time between iterations
  • Make sure that there are no spaces between the two numbers
  • Default values for i is 1, and m is 0 –> That is the default is 1 iteration and no wait time


 [-l login[:password]] (rsh/ssh login and password for rsh only)
  • Login Account to connect to Filer

-S pw:

[-S pw:|kf:]

Processing Steps:


  • The version is stated
  • In our case, the information stated is “PerfStat v7.38 (10-2012)”

Begin Iteration <Iteration>

  • Indicates the beginning of the Iteration
  • In our case, the information stated is “Begin Iteration <nth> 

Checking filer <filer>

  • Checking filer …. Establishes a connection with the filer noted
  • In our case, the information stated is “Checking filer filerHR

Prestats on <filer>; OS: ONTAP<version>

  • This step connects to the Filer and kicks off statistics gathering on the Filer
  • In our case, the information stated is “Prestats on  filerHR; OS: ONTAP8.0.2 

Sleep for <performance duration> minute(s)…

  • Once performance gathering is initiated on the Filer, this step waits for the iteration duration
  • In our case, the information stated is “Sleep for 2 minutes”

Poststats on <filer>; OS: ONTAP<version>

  • This step connects to the Filer and “concludes” statistics gathering on the Filer
  • In our case, the information stated is “Poststats on  filerHR; OS: ONTAP8.0.2 

End Iteration <Iteration>

  • Indicates the completion of the Iteration
  • In our case, the information stated is “Poststats on  End Iteration <nth> 


Sleep <n> seconds

  • This indicates how long to wait between iterations
  • In our case, the information stated is “Sleeping 60 seconds”

Sample Code (baseline):

If not exist perfData mkdir perfData

for /F "tokens=2,3,4 delims=/ " %%i in ('date/t') do set y=%%k
for /F "tokens=2,3,4 delims=/ " %%i in ('date/t') do set d=%%k%%i%%j
for /F "tokens=5-8 delims=:. " %%i in ('echo.^| time ^| find "current" ') do set t=%%i%%j
set t=%t%_
if "%t:~3,1%"=="_" set t=0%t%
set t=%t:~0,4%
rem set "theFilename=%d%%t%"
set "fname=perfData\%1__%d%%t%.perfdata"
echo %fname%

Time /T

perfstat -f %1 -t 2 -i 4,1 -l root  -S pw:rootpwd > %fname%

Time /T

Sample invokation:




    getNetAppFilerPerfData filerHR


PerfStats - Output - 20130301


NetApp – nSanity

NetApp – nSanity

nSANity Data Collector is a support tool designed to aide users and technical support in troubleshooting complex issues. nSANity is able to collect diagnostic and configuration data from a variety of components including:

To get it, you need a NetApp account.

Here are a couple of usage documentation.


    nsanity windows:[domain-name]\[user-name]:*@[host-name]

Sample - connect to local computer, using current user's credentials

   nsanity windows://localhost

Sample - use current user's credentials

   nsanity windows://dbHR

Sample (enter password in clear text)

   nsanity windows://corp\daniel:mypwd@dbHR

Sample (enter password when prompted)

   nsanity windows://corp\daniel:*@dbHR

There are some important details when you run this on a MS Windows platform and target a MS Windows host.  Here they are:

  • If you enter credentials and target your current machine, ensure that the username and password are correct; as things will not work otherwise.  Even though, this requirement must be met you will subsequently be told “User credentials can not be specified for local connections, retrying with current user credentials
  • Based on one the comments posted by a user, the password entered can not be more than 8 characters. I can confirm that as version 1.2.10 that is no longer the case.

On MS Windows Host:

On MS Windows host, please make sure that you have the following installed:

MFC90 – Requirement

The Windows executable requires MFC90 Runtime libraries, which are included with Windows 7. If your Windows host does not have the required libraries then they may be downloaded from Microsoft at the following URL.

MFC 90 is bundled with Microsoft Visual C++ 2008 SP1

FCINFO for Windows 2003

Microsoft Windows 2003 hosts require an additional package in order to allow complete data collection of HBA information. The Microsoft package is call fcinfo, which provides the HBAFAPI and WMI classes to access the API.

This package may be downloaded directly from Microsoft:


NetApp – Lun Aligning

NetApp – LUN Aligning

Determining Alignment on NetApp LUNS



   priv set diag
   lun show -m
   lun alignment show [lun-name]


    priv set diag
    lun show -m
    lun alignment show /vol/volDBBackup/q0/bkDaily


    Multiprotocol type: windows_2008
    Alignment: aligned
    Write alignment histogram percentage: 99, 0, 0, 0, 0, 0, 0, 0 
    Read alignment histogram percentage: 100, 0, 0, 0, 0, 0, 0, 0
    Partial writes percentage: 0
    Partial reads percentage: 0

In our case:

  1. We have a volume correctly “formatted” as a MS Windows 2008 Volume
  2. Based on the Alignment entry, we are Aligned
  3. The Write and Read Alignment histogram percentage has most of our non-zero entries on the left ; this indicates that data is being written and read during the initial cycles
  4. And, finally our partial writes and reads percentile are close to zero


Now how does this work? See the alignment line, that is determined from the amount of misaligned requests. Note that request can be made as well as from the OS as from applications like databases. The numbers in line after the histogram are divided in 8 equal percentages. Every number is the amount 12.5% aligned steps, from really aligned to not aligned at all. So the higher the number on the left, the better.
Sample of a MisAligned LUN:

    Multiprotocol type: windows_2008
    Alignment: misaligned
    Write alignment histogram percentage: 4, 3, 4, 5, 5, 4, 4, 3
    Read alignment histogram percentage: 6, 4, 5, 18, 11, 4, 3, 2
    Partial writes percentage: 64
    Partial reads percentage: 43

In conclusion, as a DBA you ‘re staying up late at night and during the day ‘ving Joshua Ladder’s fights and conversations with your Storage Engineers.  Next in line are calls to OS, Database, and Application Vendors.

A bit like trying to make sense of it all …

You have read Jimmy May and Denny Lee’s published papers on “Disk Partition Alignment”.   And, have taking a liking to Joe Chang’s Database & Storage poetries.

And, so now you ‘re wondering what to do.

Well like all problems worth your time, it depends:

  1. Understand your Application and read through everything the Vendor has placed in the Public Domain per I/O interaction
  2. Classify your Application – Is it doing straight through or batched writing
  3. How much pre-fetch is occurring
  4. If Database, are you really following best practice and have dedicated LUNS for Data, Log, Temporary, and Backup purpose
  5. Are your files pre-sized; and if not, do you have well orchestrated growth strategy
  6. Back to the purpose of the LUN, if SQL Server, is this the Active or Passive side of a Mirrored Pair



Addendum – 2013-04-02

For me the principal basis for quickly getting information out and broadly sharing is that you allow that information to be read and washed out in public.

Early today, received a comment via email that some of my reading about NetApp LUN Alignment is a bit wrong:
The part about the NetApp filer is wrong. The 8 figures in the histogram row are not about the percentage of overall misalignment, they are about which 512 Byte “sector” of a 4k wafl block got hit the most. Obviously you want to have 100% at the first figure, and all zeroes on the rest. Such a histogram would tell you that everything is aligned perfectly. If you have the highest value on the 7th figure (or 6th? cant remember of the top of my head) its probably a misaligned XP or windows 2003 system accessing the LUN. if you have spikes on the 1st and the 7th value, then the LUN is probably a vmware datastore, with some aligned and some misaligned VMs.
And, so here I am trying to properly credit the commenter and see if there is new information on the Net along same lines.

And, as always Google brought us joy.

LUN (mis) alignment:

lun (mis)alignment
‎2011-07-14 06:47 AM – edited ‎2015-12-18 01:25 AM
LABELS: SnapManager




   priv set -q advanced;
   lun show -v


    priv set -q advanced;
    lun show -v

NetApp - Lun Show -v

On this new Filer, returned to running “lun alignment show”:



   priv set advanced
   lun alignment show 


    priv set advanced
    lun alignment show -m

NetApp - Lun alignment show


  • The “partial writes percentage” and “partial reads percentage” numbers returned from “lun alignment show” is congruent with the summative data from “lun show -v”; specifically from “Alignment entry” –> misaligned /aligned.
  • In our case, the two LUNS that indicate “misaligned” are the luns dedicated to SQL Server database logs.

On the Windows Host, to determine the “bytes per cluster” for a logical drive, issue:


    fsutil fsinfo ntfsinfo [logical-drive]:


    -- get bytes per cluster for logical drive L:
    fsutil fsinfo ntfsinfo L:

MS - Windows - Drive - Bytes per cluster

Again for Microsoft OS, you want to see “Bytes per cluster” coming in as 65536.
On the Windows Host, to determine the Partition “StartingOffset” for a logical drive, issue:


    wmic partition get BlockSize, StartingOffset, Name

WMIC Partition


Based on NetApp Published guidance, this number can be one any of 3 numbers depending on the Version of Windows and the Disk Format Style:

How to diagnose misaligned I/O on Windows hosts

  • Windows (MBR): This number should be 32256, 31.5kb offset is used when the LUN is created with the Windows LUN type (31.5 * 1024 = 32256).
  • Windows (GPT): This number should be either 33,571,840B (32MB+17KB if LUN is <16GB) or  134,235,136 (128MB + 17KB if LUN >=16GB).
  • Windows (2000): This number should be 65535 bytes for LUNs smaller than 4GB or 1048576 bytes for LUNs that are 4GB or larger.

Our Starting Offset is 1048576 and so it matches what we should per the NetApp KB listed above.

Addendum – 2013-05-16

Here is mostly the same information, but looked at with more understanding.

It is a silly thing, but like Snoop said:
… You don’t think of all of that, until you talk about all of that


  1. Lun Aligning
  2. New VDMK Misalignment Detection Tools in dataontap 725 and 802
  3. LUN (Mis)alignment
  4. How fragmentation on incorrectly formatted NTFS Volumes affects Exchange
  5. How to diagnose misaligned I/O on Windows hosts