Windows Server Setup

Information for setting up Windows Server on a new Server

Introduction

CF Foundry uses an array of Windows Server machines that serve as Web Servers, Application Servers, and/or Database Servers. This part of our documentation outlines how to setup these various machines and related software applications.

Supporting Large Drive Arrays

Our servers are no generally supporting RAID 10 arrays that use hard drives of several gigabytes. To avoid the 2TB drive restrictions of standard BIOS on these servers, the servers are configured to use UEFI (Unified Extensible Firmware Interface). In order to use UEFI on a server, you will need to:

Ensure that the server has a RAID controller that will support larger drives. For example, on DELL servers, the H700 will support single drives larger than 2TB, whereas the PERC6 will not.
Format a USB drive so it will be bootable by a UEFI system. This means changing the partition format of the partition to a GPT (not MBR), the file system to FAT32, and limiting the partition size to 16GB.
Go into the server BIOS (F2 during the boot cycle for DELL Servers) and switch the boot control from BIOS to UEFI. Note that, on some older servers, this may require updating firmware.
Insert the USB stick into one of the USB slots.
Set the boot order to include the USB ports. Ensure that the USB ports are not disabled in the BIOS for security purposes.
Reboot the server and load the OS from the USB.

Creating a Bootable GPT USB Drive

To create a USB stick that can be booted by UEFI, follow these instructions:

Open an elevated command prompt
Insert the USB stick
Type diskpart and hit enter
Type in the following commands:

List disk
Sel disk X [where x: is the number of the USB drive]
Clean
convert gpt
Create Part Primary size=16000 [use size only if the USB is > 16GB]
Format fs=fat32 quick label="Boot"
Assign letter="k" [use an any available drive letter]
exit

At this point you can now copy you boot files to the FAT32 partition. Mount the ISO image by double-clicking on it in windows. This will assign it a drive letter and allow you to copy the file contained within to the new USB drive.

NOTE: One of the files for the server installation will likely be over 4GB, which is too large for a FAT32 file system. To get around this, use the following command:

dism /Split-Image /ImageFile:e:\sources\install.wim /SWMFile:k:\sources\install.swm /FileSize:4096

This assumes that your ISO image is mounted as drive "E" and the USB as drive "K". Otherwise, substitute your correct drive letters.

An alternative to this process is to use the Rufus application as follows:

Start the tool as administrator
Select the USB storage device then make the partition scheme and target system type to be GPT “partition scheme for UEFI”.
Give it the ISO image of windows server 2016 or whichever O.S. you are installing and let the tool create the bootable USB.
In the server boot settings make it UEFI and then make the partition GPT.

Creating Bootable Windows Server 2016 USB

Download ISO image from Microsoft's Volume Licensing Service Center (VLSC).
Double-click on the ISO to mount it.
Insert the USB drive into the PC and make note of the assigned drive letter.
Open an elevated command prompt (right-click on cmd and chose to run as Administrator).
Run "DiskPart" (disk partition utility) as shown below in figure 1. Wherever the "?" appears, replace it with the drive letter of the USB.
You should be exited out of DiskPart but still in the elevated command window. Type in the drive letter of the USB to move the local root to the USB and then change to the "boot" directory on the drive by typing, 'cd boot'.
Load the Boot Manager code onto the USB by running "BootSect" by typing, 'bootsect /nt60 e:' in the elevated command prompt. Append the '/force' switch if needed.
Run the xCopy command to copy all file. Again, replace the "?" with your drive letter for the USB and replace the X with the drive letter of the mounted ISO image. Note that it will take a considerable amount of time to copy the install.wim file as this is the primary image file and is several gigabytes in size.
xcopy x:\*.* ?:\ /E /H /F
When loading the USB into the server, be sure to boot from the USB drive by selecting F12 (Dell Servers).
If you need to prep the drive array for RAID services (we typically use RAID 10), do this before booting the USB by pressing Ctrl-R during the BIOS boot to enter the PERC or H series RAID controller BIOS setup when prompted (again, for DELL Servers).

Figure 1

diskpart
list disk
select disk ?
list disk
— Note: after you run list disk again, the drive should be prefaced with an asterisks “*”
clean
create partition primary
select partition 1
active
format fs=ntfs quick label=”2016SRV”
exit
    Leaving DiskPart…

PreviousStatusPage Integration NextCommon CF Formatting Commands

Last updated 5 years ago

Was this helpful?

Supporting Large Drive Arrays

Ensure that the server has a RAID controller that will support larger drives. For example, on DELL servers, the H700 will support single drives larger than 2TB, whereas the PERC6 will not.

Format a USB drive so it will be bootable by a UEFI system. This means changing the partition format of the partition to a GPT (not MBR), the file system to FAT32, and limiting the partition size to 16GB.

Go into the server BIOS (F2 during the boot cycle for DELL Servers) and switch the boot control from BIOS to UEFI. Note that, on some older servers, this may require updating firmware.

Insert the USB stick into one of the USB slots.

Set the boot order to include the USB ports. Ensure that the USB ports are not disabled in the BIOS for security purposes.

Reboot the server and load the OS from the USB.

Creating a Bootable GPT USB Drive

To create a USB stick that can be booted by UEFI, follow these instructions:

Open an elevated command prompt

Insert the USB stick

Type diskpart and hit enter

Type in the following commands:

List disk
Sel disk X [where x: is the number of the USB drive]
Clean
convert gpt
Create Part Primary size=16000 [use size only if the USB is > 16GB]
Format fs=fat32 quick label="Boot"
Assign letter="k" [use an any available drive letter]
exit

NOTE: One of the files for the server installation will likely be over 4GB, which is too large for a FAT32 file system. To get around this, use the following command:

dism /Split-Image /ImageFile:e:\sources\install.wim /SWMFile:k:\sources\install.swm /FileSize:4096

This assumes that your ISO image is mounted as drive "E" and the USB as drive "K". Otherwise, substitute your correct drive letters.

An alternative to this process is to use the Rufus application as follows:

Format your USB stick to UEFI and GPT by downloading the tool from:

Start the tool as administrator

Select the USB storage device then make the partition scheme and target system type to be GPT “partition scheme for UEFI”.

Give it the ISO image of windows server 2016 or whichever O.S. you are installing and let the tool create the bootable USB.

In the server boot settings make it UEFI and then make the partition GPT.

Creating Bootable Windows Server 2016 USB

Download ISO image from Microsoft's Volume Licensing Service Center (VLSC).

Double-click on the ISO to mount it.

Insert the USB drive into the PC and make note of the assigned drive letter.

Open an elevated command prompt (right-click on cmd and chose to run as Administrator).

Run "DiskPart" (disk partition utility) as shown below in figure 1. Wherever the "?" appears, replace it with the drive letter of the USB.

You should be exited out of DiskPart but still in the elevated command window. Type in the drive letter of the USB to move the local root to the USB and then change to the "boot" directory on the drive by typing, 'cd boot'.

Load the Boot Manager code onto the USB by running "BootSect" by typing, 'bootsect /nt60 e:' in the elevated command prompt. Append the '/force' switch if needed.

Run the xCopy command to copy all file. Again, replace the "?" with your drive letter for the USB and replace the X with the drive letter of the mounted ISO image. Note that it will take a considerable amount of time to copy the install.wim file as this is the primary image file and is several gigabytes in size.

xcopy x:\*.* ?:\ /E /H /F

When loading the USB into the server, be sure to boot from the USB drive by selecting F12 (Dell Servers).

If you need to prep the drive array for RAID services (we typically use RAID 10), do this before booting the USB by pressing Ctrl-R during the BIOS boot to enter the PERC or H series RAID controller BIOS setup when prompted (again, for DELL Servers).

Figure 1

diskpart
list disk
select disk ?
list disk
— Note: after you run list disk again, the drive should be prefaced with an asterisks “*”
clean
create partition primary
select partition 1
active
format fs=ntfs quick label=”2016SRV”
exit
    Leaving DiskPart…