7. Postinstallation tasks
After you’ve installed a server and logged on, you might be ready to
call it a day. Don’t do this yet because you should first perform a few
final postinstallation procedures. The Tools menu in Server Manager
gives you quick access to tools for administration. Using Server Manager
as your starting point, you can do the following:
-
Check devices
Select Device Manager on the Tools menu, and then use Device Manager,
to look for undetected or malfunctioning hardware components. If you
find problems, you might need to download and install updated drivers
for the computer—you can download from another system and then transfer
the files to the new server using a USB key or by burning the files to a
CD/DVD-ROM. If you removed any system hardware prior to installation,
you might want to add it back in and then check again for conflicts and issues that must be corrected. You aren’t finished with Device Manager until every piece of hardware is working properly.
-
Check the TCP/IP configuration
When you select the
Local Server node in Server Manager, you’ll see the server’s basic
configuration settings. Tap or click the links for the Ethernet settings
to open the Network Connections dialog box. Ensure that the TCP/IP
configuration is correct and that any additional settings are applied as
necessary for the network. Test TCP/IP networking from the command line
using Ping or Tracert and in the Windows operating system by trying to
browse the network.
-
Check event logs When you select the Dashboard node, you’ll see an entry for the Local Server. If there are errors
or critical events, you can tap or click Events to review them. You
also can use the Events panel under Local Server and the Event Viewer to
check the Windows event logs. Any startup warnings or errors will be
written to the logs.
-
Check disk partitioning
Select Computer Management on the Tools menu, and then use the Computer
Management console to check and finalize the disk partitions. Often,
you must create the server’s application partition or configure software
RAID.
-
Optimize system configuration ,
for tuning the operating system. For example, you might need to change
the display settings, virtual memory pagefile usage, or the Server
service configuration. You might also need to add local group and user
accounts to the server per standard IT procedures.
-
Update the server
Use Windows Update or Windows Server Update Services to ensure that the
operating system is up to date and has the most recent updates for
stability and security. When Windows Update is configured properly, you
can tap or click the Windows Update link on the Local Server node and
then tap or click Check For Updates to get updates for the server.
-
Reboot for good measure After
you configure the server and optimize its settings, perform a final
reboot to ensure that (1) the server starts, (2) all the server services
start, and (3) no other errors occur. You should reboot even if the
changes you made don’t require it—it’s better to find out about problems
now rather than at 3 A.M. on a Saturday night.
-
Prepare backup and recovery You’re almost done, but not quite. Don’t forget about creating an automated
recovery disk for the server. You might also want to perform a full
backup.
These postinstallation
procedures are not only important, they’re essential to ensuring that
the server performs as well as can be expected. After these procedures
are completed, you should have a server that is nearly ready for its
role in a production environment. Don’t make the server available to
users just yet. To finish the job, you need to install and configure any
necessary roles, role services, features, and applications. For
certain, configuring these components requires quite a bit of extra work
beyond installing the operating system. The installation
of these additional components and applications could require one or
more reboots or might require several time periods in which users are
blocked from accessing the server or in which they are requested not to
connect to it as well. Remember, from the users’ perspective, it’s
usually better to not have a resource than to be given one and then have
it taken away (even temporarily). Finalize the server, and then deploy
it, and you’ll have happier users.
You can convert the installation
type. To convert a Full Server installation to a Minimal Server
Interface installation, you remove the Server Graphical Shell. Although
you can use the Remove Roles And Features Wizard to do this, you also
can do this at a Windows PowerShell prompt by typing the following
command:
uninstall-windowsfeature server-gui-shell -restart
This command instructs Windows Server to uninstall the Server
Graphical Shell and restart the server to finalize the removal. If
Desktop Experience also is installed, this feature will be removed as
well.
To convert a Minimal Server Interface installation to a Server With A
GUI installation, you add the Server Graphical Shell. You can use the
Add Roles And Features Wizard to do this, or you can type the following
command at a PowerShell prompt:
install-windowsfeature server-gui-shell -restart
This command instructs Windows Server to install the Server Graphical
Shell and restart the server to finalize the installation. If you also
want to install the Desktop Experience, you can use this command
instead:
install-windowsfeature server-gui-shell, desktop-experience -restart
To convert a Full Server or Minimal Server Interface installation to a Server Core
installation, you remove the user interfaces for Graphical Management
Tools And Infrastructure. If you remove the WOW64 Support framework, you
also convert the server to a Server Core installation. Although you can
use the Remove Roles And Features Wizard to remove the user interfaces,
you also can do this at a PowerShell prompt by typing the following
command:
uninstall-windowsfeature server-gui-mgmt-infra -restart
This command instructs Windows Server to uninstall the user
interfaces for Graphical Management Tools And Infrastructure and restart
the server to finalize the removal. Because many dependent roles, role
services, and features might be uninstalled along with the user
interfaces, run the command with the –Whatif parameter first to get details on what exactly will be uninstalled.
If you installed the server with the user interfaces and converted it
to a Server Core installation, you can revert back to a Full Server
installation with the following command:
install-windowsfeature server-gui-mgmt-infra -restart
As long as the binaries for this feature and any dependent features
haven’t been removed, the command should succeed. If the binaries were
removed, however, or Server Core was the original installation type, you
need to specify a source for the required binaries. If you don’t do
this, the required feature will be downloaded from Windows Update, which
could take a long time for some features.
You use the –Source
parameter to restore required binaries from a Windows Imaging (WIM)
mount point. For example, if your enterprise has a mounted Windows Image
for the edition of Windows Server 2012 you are working with available
at the network path \\ImServer18\WinS12EE, you could specify the source
as follows:
install-windowsfeature server-gui-mgmt-infra -source \\imserver18\wins12ee
Although many large enterprises might have standard images that can
be mounted using network paths, you also can mount the Windows Server
2012 distribution media and then use the Windows\WinSXS folder from the installation image as your source. To do this, follow these steps:
-
Insert the installation disc into the server’s disc drive, and then
create a folder to mount the Installation image by typing the following
command: mkdir c:\mountdir.
-
Locate the index number of the image you want to use by typing the following command at an elevated prompt: dism /get-wiminfo /wimfile:e:\sources\install.wim, where e: is the drive designator of the server’s disc drive.
-
Mount the installation image by typing the following command at an elevated prompt: dism /mount-wim /wimfile:e:\sources\install.wim /index:2 /mountdir:c:\mountdir /readonly, where e: is the drive designator of the server’s disc drive, 2 is the index of the image to use, and c:\mountdir is the mount directory. Mounting the image might take several minutes.
-
Use Install-WindowsFeature at a PowerShell prompt with the source specified as c:\mountdir\windows\winsxs, as shown in this example:
install-windowsfeature server-gui-mgmt-infra
-source c:\mountdir\windows\winsxs
