Transcript Module 6: Modifying Cluster Resources

Module 6: Creating
Cluster Resources
Overview

Creating a File Share Resource

Creating a Cluster Print Share

Configuring Clustered Applications and Services

Using Cluster Application Wizard

In this module you will learn how to add resources,
such as a file or print shares, and configure generic
applications and services. You will be using the Cluster
Administrator tool to perform the tasks instead of the
Cluster Application Wizard. However, you can perform
the same tasks by using the Cluster Application Wizard.
After completing this module, you will be able to:

Create a file share resource.

Create a cluster print share.

Configure generic applications and services.

Launch the Cluster Application Wizard.

Note: For more information about how to use the
Cluster Administrator tool to create cluster resources,
see Module 5, “Cluster Administration,” in Course
2087A.
 Creating a File Share Resource
Cluster Administrator – [CLUSTER (server1)]
File View Window Help
CLUSTER
Groups
Name
State
Cluster IP Address Online
er Name
Online
Online
Resources Bring Online
Cluster Co Take Offline
Resour Move Group
Netwo
Delete
Clu
Rename
Pul
Netwo New
SERVER1 Configure Application
Active
Active Properties
Network Interfaces
SERVER2
Active Groups
Active Resources
Network Interfaces
Creates a new cluster group
Owner
DC1
DC1
DC1
Resource Type
IP Address
Network Name
Physical Disk
New Resource
Project files share
Possible Owners
Group
Resource
Name:
Users
Description:
Project files share
Possible
owners
are nodes
in the Disk
cluster on which this resource can be brought online.
Resource
type:
Physical
Specify the possible owners for this resource.
Group:
DHCP Service
Distributed Transaction CoordinPossible owners:
Available nodes:
File Share
Run this resourc Generic Application
Name
Name
Generic Service
SERVER1
To continue, click N IIS Server InstanceAdd ->
SERVER2
IP Address
Message Queuing
<- Remove
Network Name
NNTP Server Instance
Physical Disk
< Back
Next >
Cancel
< Back
Next >
Cancel

After creating file shares on a cluster, you can configure
them to be fault tolerant. You can use these shares for
normal file shares, home folders, and the Distributed
File System (DFS) root. There are three major ways that
you can use this file share resource type:




Use a basic file share for sharing a single folder.
Use share subdirectories to have Cluster service create
large numbers of related shares automatically.
Create a DFS root to provide fault tolerance for a standalone DFS root.
Create a file share resource, which is similar to creating
other resources by using Cluster Administrator.

From Cluster Administrator, select New, and then click
Resource.

Enter a name and description of the resource for
administrative purposes. Select File Share, as the
Resource type. Identify the group in the cluster to which
this resource will belong.

Select the Possible owners for failover purposes.

You will next establish dependencies and set access
permissions and file share settings.
File Share Dependencies
Cluster Administrator – [CLUSTER (server1)]
File View Window Help
Name
State
Owner
Resource Type
Cluster IP Address Online
DC1
IP Address
er Name
Online
DC1
Network Name
New Resource
Resources Bring Online
Online
DC1
Physical Disk
Cluster Co Take Offline
Possible Owners
Resour Move Group
Project files share
Netwo
Delete
Clu
Users
Dependencies Project files share
Name:
Rename
Pul
Possible owners
nodes in the cluster on which this resource can be brought online.
Netwo New
Description:
Group are
Specify the
possible
owners
for this resource.
SERVER1 Configure Application
Resource
Resource type:Users Physical Disk
Active
Available nodes:
Possible owners:
Group:
DHCP Service
Active Properties
Name
Dependencies are
resources
whichName
must
be brought online by the cluster service first.
Distributed
Transaction
Coordin
Network Interfaces
File Share for this resource.
Specify the dependencies
SERVER1
Run this resourc Generic
SERVER2
AddApplication
->
SERVER2
Available resources:
Resource dependencies:
Generic Service
Active Groups
IIS
Server
Instance
To
continue,
click
N
Active Resources
<- Remove Reso
Resource
Resource
Reso
IP Address
Network Interfaces
Cluster IPMessage
AddressQueuing
IP Ad
Cluster Name
Netw
Add ->
CLUSTER
Groups
Creates a new cluster group
Disk G: Network Name Phys
NNTP Server Instance
Physical Disk
< Back
<- Remove
Next >
Cancel
< Back
Next >
< Back
Cancel
Next >
Cancel

When you set up a file share resource in Cluster
Administrator, you are not required to enter any
dependencies for the file share. However, it is
recommended that a file share resource depend on at least
a network name resource and a physical disk resource.

The network name dependency enables the client to access
the file share over the network by using the virtual server
name instead of the node name. Using the virtual server
name ensures that the client opens the connection to
whichever node is the current owner of the file share
resource. The physical disk resource dependency ensures
that clients have access to the files on the disk through the
share.

It is recommended that the share name be dependent on
the virtual server name, the server name be dependent on
the Internet Protocol (IP) address, and the IP address
dependent on the disk. These dependencies create a
vertical dependency tree, which is useful for troubleshooting,
and also provide consistency for the resources in the cluster.

Note: The file share resource does not place a mandatory
requirement on a physical disk resource that corresponds to
a device on the shared bus. Not requiring a mandatory
dependency allows you to use other forms of storage for
your file share resource under special circumstances. These
alternative forms of storage may include local storage for
nonvolatile data in which the same data exists on both
nodes.
File Share Parameters
Cluster Administrator – [CLUSTER (server1)]
File View Window Help
CLUSTER
Groups
Permissions…
Security
Name
State
Cluster IP Address Online
er Name
Online
Online
Resources Bring Online
Cluster Co Take Offline
Move Group
Resour
File Share
Parameters
Netwo
Delete
Clu
Rename
Pul
Users
Netwo New
SERVER1 Configure Application
ShareActive
name:
Users
Active Properties
Path:Network Interfaces
g:\users
SERVER2
Comment:
Users Home Folders
Active Groups
Active
User
LimitResources
Network Interfaces
Maximum allowed
Allow
Creates a new cluster group
Permissions for Public (Public)
Owner Name
Resource Type
Everyone IP Address
DC1
DC1
Network Name
DC1
Physical Disk
Add…
Remove
Permissions:
Group
Resource
Allow
Deny
Full Control
Change
Read
Advanced File Share Properties
users
Normal share
Dfs root
Advanced…
Share subdirectories
Hide subdirectory shares
< Back
Finish
Cancel
OK
Cancel

Every resource has a resource name, possible owners, and
dependencies, but depending on the resource type, you will
need to configure a different set of parameters. File share
parameters on a virtual server have the same configurations
as a file share on a physical server, with the exception of the
settings for the Advanced File Share Properties dialog
box.

The table below shows the specific File Share Parameter
options.
Option
Description
Share name
Unique name for the share.
Path
Comment
Path for the shared folder, on a shared drive. The folder
on the shared drive must already exist.
Description for the share (optional).
User Limit
Maximum number of concurrent users.
Permissions
Configuration of the file share permissions.
Advanced
Specify whether subdirectories of the shared folder
should be available through the single share resource.
Permissions for the File Share

Use Cluster Administrator and Microsoft Windows 2000
security to limit access to files and folders that reside
on drives on the cluster storage device.

When setting access permission for file shares, it is
recommended that you base file share permissions on domain
local groups, not on local user or local group accounts. On
member servers, local users and local groups have a security
context only on the local computer. The security context of
these accounts and groups is meaningless when failed from
one node to another. For this reason, you cannot give local
users or local groups permissions to administer the cluster.
The single exception to this rule is the local Administrator
group. The local Administrator group is not a problem on
domain controllers because the local accounts and groups
have security context on all domain controllers in the domain.
It is recommended that you also set NTFS permissions through
domain local groups.

Important: If you change file share permissions by using
Windows Explorer or My Computer instead of using
Permissions on the Parameters tab in Cluster Administrator,
the permissions are lost when the resource is taken offline.
Advanced File Share Properties

A normal file share is the default option. If you want to
share subdirectories or create a DFS root, click the
Advanced button. There are three options: Normal
Share, Dfs root, and Share subdirectories (dynamic
shares).
DFS Root

A DFS root is the highest level of a DFS topology, and is
the starting point for the hierarchy of shared folders. A
domain DFS root can have replicas for fault tolerance,
but a stand-alone DFS root is lost if the server hosting
the DFS root goes offline. Cluster service can provide a
fault tolerant DFS root by placing the DFS root on a
virtual server. The DFS root that Cluster service
configures is a stand-alone DFS root, not a domain DFS
root.

The DFS root that uses Active Directory™ directory
service supports automatic file replication. Stand-alone
DFS root configurations are not automatically replicated
outside of the server cluster. The server cluster
provides failover of the DFS root for availability rather
than relying on replicas of the DFS topology.

Caution: If there is an existing DFS root that you have
configured on any node in the server cluster, you must
either remove it or migrate it to the cluster.
Share Subdirectories (Dynamic Shares)

Every cluster resource has a performance impact;
Cluster service must periodically poll all of the
resources to determine if they have failed. Because of
this poll, you may notice that your cluster’s
performance decreases as you increase the number of
resources that it manages. The worst-case scenario is
that a very busy cluster may drop connections because
of the amount of time that is being used to manage a
high volume of cluster resources. Large quantities of
file share resources also increase the length of the
failover time.

The optimal way to use a server cluster for home folders
is to use the Share subdirectories (dynamic shares)
feature. Cluster service allows you to configure multiple
shares as a single resource. The resource parameters
permit you to share subdirectories separately from the
parent.

Furthermore, it is possible to configure the resource so
that you create the subdirectory shares as hidden
shares that do not appear in browse lists. This feature
prevents the need to use multiple file share resources to
create large numbers of shares, which in turn
dramatically reduces overhead and simplifies
manageability of large numbers of shares.
Lab A: Creating Cluster File Shares
Objectives
After completing this lab, you will be able to:

Create a file share resource.

Configure a dynamic share.

Create home folders without dynamic shares.

Create home folders with dynamic shares.
Prerequisites

Before working on this lab, you must be familiar with the
concepts in the section Creating a File Share Resource
in Module 6, “Creating Cluster Resources,” of Course
2087A. You must also have a server cluster that is
installed and running on both nodes.
Lab Setup
To complete this lab, you need the following:

Two computers running Microsoft Windows 2000
Advanced Server, each with a small computer system
interface (SCSI) adapter installed. Each computer in a
student pair must have a different SCSI ID.

A shared external SCSI disk. Each computer is
connected to the shared disk with SCSI cables. The
shared external SCSI disk ID must be different than
either of the SCSI cards of the computers in the student
pair.

Two network adapter cards in each node of the cluster.

The following files: usrfldr.bat, makeshare.vbs, and
removeshare.vbs
Scenario

In this exercise, you have been asked to create a file share for a
public folder on a virtual server. You are then to configure this file
share resource as a dynamic share. You need to create file share
resources for 50 user home folders and test the failover time of
these resources. You will remove the 50 user home folders and
create one dynamic share for 200 user home folders and test the
failover time of this resource.

The following exercises will refer to your computers as Node A and
Node B. For Exercise 1, you will take turns creating different parts
of a file share resource. In Exercise 2, you will take turns
configuring a dynamic share from the file share resource that is
created in Exercise 1. In Exercise 3, you will take turns creating
multiple file share resources that are used for user home folders
and converting these shared resources into dynamic shares. You
will work together to time the failover of the file share resources.
Exercise 1: Creating a File Share Resource

In this exercise you will create a file share for a virtual
server.
To create a folder on the shared disk
Complete this task from Node A.
1. Click Start, point to Programs, point to Accessories, and
then click Windows Explorer.
2. In Windows Explorer, expand My Computer, and then
click Shared Disk 1 (W:).
3. On the File menu, point to New, and then click Folder.
4. Type Public to create the folder name.
5. Close Windows Explorer.
To create a file share resource
Complete this task from Node B.
1. Click Start, point to Programs, point to Administrative Tools,
and then click Cluster Administrator.
2. In the console tree, expand Cluster Name, and then expand
Groups.
3. Right-click Cluster Group, point to New, and then click
Resource.
4. Use the following information to complete the New Resource
dialog box: Name: Public Share Description: Public Share
Resource Type: File Share Group: Cluster Group
5. Click Next.
6. In the Possible Owners dialog box, click Next.
7. In the Dependencies dialog box, under Available
resources:, select Disk W:, and then click Add.
8. Click Next.
9. Use the following information to complete the File
Share Parameters dialog box: Share Name: Public
Path: W:\Public Comment: Public Share on W:
10. Click Finish.
11. In the Cluster Administrator dialog box, click OK.
12. In Cluster Administrator, right-click Cluster Group, and
then click Bring Online.
To test a file share resource
Complete this task from Node A.
1. Click Start, and then click Run.
2. In the Run dialog box, type \\Virtual_Server\Public and
then click OK.
3. Close Windows Explorer.
Exercise 2: Configuring a Dynamic Share

In this exercise you will create additional folders in the
Public folder. The dynamic share feature of Cluster
service will share out the newly-created Public folders
on the virtual server.
To create a dynamic share
Complete this task from Node A.
1. Click Start, point to Programs, point to Administrative Tools, and
then click Cluster Administrator.
2. Expand Groups, select Cluster Group, and in the details pane rightclick Public Share, and then click Properties.
3. In the Public Share Properties dialog box, click the Parameters tab.
4. Click Advanced.
5. In the Advanced File Share Properties dialog box, click Share
subdirectories, and then click OK.
6. Click OK to close the Public Share Properties dialog box.
7. Click OK to accept the Cluster Administrator Standard Extension
error message.
8. Right-click Public Share, and then click Take Offline.
9. Right-click Public Share, and then click Bring Online.
To create folders on the Public Share
Complete this task from Node B.
1. On the Start menu, click Run.
2. In the Run dialog box, type \\Virtual_Server\Public
3. From the Windows Explorer window, click File, select
New, and then select Folder to create the following
folders; repeat this step each time until the three folders
have been created.
Engineering
Sales
Marketing
To test the dynamic share on the Public Share
Complete this task from Node A and Node B.
1. On the Start menu, click Run.
2. In the Run dialog box, type \\Virtual_Server and click
OK.
3. Verify that in the Windows Explorer windows you have
three additional shares named Engineering, Sales, and
Marketing.
Exercise 3: Configuring User Home Folders

In this exercise you will use a batch file to create 200
folders to be used for home folders on your W: drive.
You will then run another script to create a file share
resource for 50 of the 200 home folders. You will test the
failover time of the 50 file share resources and run
another script to remove the 50 file share resources.
You will then configure a file share resource for dynamic
shares to provide 200 user home folders and test the
failover time of the dynamic share feature of Cluster
service.
To create 200 user folders on the shared disk
Complete this task from Node B.
1. Click Start, point to Programs, point to Administrative Tools, and then click
Cluster Administrator.
2. Expand Groups, and then select Cluster Group.
3. In the details pane, verify that the owner is Node B. If the Owner is not
Node B, right-click Cluster Group, and then click Move Group.
4. Click Start, and then click Run.
5. In the Run dialog box, type c:\moc\2087a\labfiles\mscs\usrfldr.bat and then
click OK.
6. To verify that the batch file ran successfully, in the Run dialog box, type
w:\users and then click OK
7. Verify that there are 200 folders that were created in the users folder. You
should see user1 through user200.
8. Close Windows Explorer.
9. Leave Cluster Administrator running.
To create 50 file share resources in the cluster
Complete this task from Node A.
1. In the Run dialog box, type
c:\moc\2087a\labfiles\mscs\makeshare.vbs and then
click OK
2. Wait until a message box appears saying that the file
creation is complete, and then click OK.
To move 50 file share resources from one node to the
other
Complete this task from Node B.
1. Write down the time from the system tray at the bottom
right corner of your desktop._______________
2. In Cluster Administrator, right-click Cluster Group, and
click Move Group. This step could take more than three
minutes to complete.
3. Write down the time from the system tray at the bottom
right corner of your desktop._______________
4. How long did it take to move the 50 file share
resources?_________
To remove the 50 file share resources from the
cluster
Complete this task from Node A.
1. In the Run dialog box, type
c:\moc\2087a\labfiles\mscs\removeshare.vbs and then
click OK.
2. When the message dialog box appears that says that all
of the users file shares have been successfully
removed, click OK.
To create a dynamic share for 200 user home folders
Complete this task from Node A.
1. Click Start, point to Programs, point to Administrative Tools,
and then click Cluster Administrator.
2. In the console tree, expand Cluster Name, and then expand
Groups.
3. Right-click Cluster Group, point to New, and then click
Resource.
4. Use the following information to complete the New Resource
dialog box: Name: Users Share Description: Users Share
Resource Type: File Share Group: Cluster Group
5. Click Next.
6. In the Possible Owners dialog box, click Next.
7. In the Dependencies dialog box, under Available resources:, select
Disk W:, then click Add.
8. Click Next.
9.
Use the following information to complete the File Share
Parameters dialog box: Share Name: Users Path: W:\Users
Comment: Users Share on W:
10. Click Advanced.
11. In the Advanced File Share Properties dialog box, click Share
subdirectories, and then click OK.
12. Click Finish.
13. In the Cluster Administrator dialog box, click OK.
14. In the Cluster Administrator dialog box, right-click Cluster Group,
and then click Bring Online.
To verify that the dynamic shares were created
Complete this task from Node B.
1. Click Start, and then click Run.
2. In the Run dialog box, type \\Virtual_Server and then
click OK.
Notice that Cluster service created a file share for every
user in the w:\users folder.
3. Close Windows Explorer.
To move 200 file share resources from one node to
the other node
Complete this task from Node B.
1. Write down the time from the system tray at the bottom
right corner of your desktop._______________
2. From Cluster Administrator, right-click Cluster Group,
and then click Move Group.
3. Write down the time from the system tray at the bottom
right corner of your desktop._______________
4. How long did it take to move the 200 dynamic shares
from one file share resource?______________
 Creating a Cluster Print Share

Installing Printer Drivers

Creating a Print Spooler Resource

Adding a Print Share

Cluster service supports the clustering of print services
by using the print spooler resource. A cluster can have
multiple print spooler resources, but each group can
only have one print spooler resource. Clients access
clustered printers in exactly the same way that they
access nonclustered printers. Clients can use either the
network name or Internet Protocol (IP) address as the
server name.

Creating a print spooler on a cluster is a more involved
process than creating a file share. A cluster print
spooler uses a network printing device or remote print
server, not a locally attached device, because any
peripherals that are local to one node will not be
available if that node is offline.

One method for creating a print spooler on a cluster requires that
you install the printer drivers on both nodes before installing them
on the virtual server. The steps are as follows:
1.
Install printer drivers on each node.
2.
Create a print spooler resource.
3.
Add a print share.

Important: An alternate method for creating a print spooler on a
cluster is described in the Q Article Q228904: Print Spooler
Support on Windows 2000 Server Cluster, on the Student compact
disk. In this method you configure the port, the drivers, and the
printer share name on the virtual server, fail over the spooler
resource to the other node, and reconfigure the drivers and print
share on the virtual server. The port will already exist on the virtual
server.
Installing Printer Drivers
Printers
File
Edit
View Favorites
Server Properties
Create Shortcut
Tools
Help
Search
Folders
History
Go
Delete
Print Server Properties
Rename
Properties
Forms
Close
Ports
Drivers
Advanced
Add Printer
HP LaserJet
4050
Series PS
MOCMOC-DC1
This folder contains information
about printers that are currently Installed printer drivers:
installed, and a wizard to help you
Name
install new printers.
Environment Version
Add Printer Driver Wizard
To get information about a printer
that is currently installed, right-click
the printer’s icon.
Welcome to the Add Printer
Driver Wizard
To install a new printer, click the
Add Printer icon.
This wizard helps you install printer drivers for various
platforms on a print server.
After the printer drivers are installed on the print server, the
appropriate printer drivers are automatically downloaded to
clients when they connect to the printer server.
Displays properties for this server.
To continue, click Next.
Add…
Remove
Update
OK
Cancel
Properties…
Apply
< Back
Next >
Cancel

You must install printer drivers on each cluster node
that will be a potential owner of a spooler resource.
When you install the printer driver on a node, the printer
driver is copied to the Print$ share. Each node can use
the printer driver to print to the print device and supply
clients with the driver if requested.

You can publish printers that are hosted on a server
cluster running Windows 2000 to Active Directory. This
option is selected by default and the printer is published
under the node that owned the print spooler resource
when you created the printer.

To install a print driver on a node of the cluster, you
need to configure the print server properties. From the
Printers icon in the Control Panel, select Server
Properties from the File menu. Click the Drivers tab,
click Add, and select the appropriate driver for the
printer that you are configuring.

Note: Each node in a clustered print server running
Windows 2000 can perform print shares outside of the
cluster; however, the print share name must be different
from all of the print share names in the cluster.
Creating a Print Spooler Resource
Cluster Administrator – [CLUSTER (server1)]
File View Window Help
CLUSTER
Groups
Name
State
Cluster IP Address Online
er Name
Online
New Resource
Online
Possible Owners
Resources Bring Online
Cluster Co Take Offline
Resour Move Group
Netwo
Delete
Clu
Rename
Pul
Netwo New
SERVER1 Configure Application
Active
Active Properties
Network Interfaces
SERVER2
Active Groups
Active Resources
Network Interfaces
Creates a new cluster group
Owner
DC1
DC1
DC1
Resource Type
IP Address
Network Name
Physical Disk
Project files share
Dependencies
Users
Name:
Project files share
Print
Spooler
nodes
in the Parameters
cluster on which this resource can be brought online.
Description:
GroupPossible owners are
Users
Specify the possible owners for this resource.
Resource
Resource type:
Physical Disk
Available nodes:
Possibleonline
owners:
Dependencies are resources which must be brought
by the cluster service first.
Group:
DHCP Service
Print Spooler
forDistributed
this resource.
Name Specify the dependencies
Name
Transaction Coordin
File Share
Available resources:
Resource dependencies:
Run this resourc Generic
AddApplication
->G:\Spool SERVER1
Spool
folder:
SERVER2
Resource
Reso
Resource
Generic Service
IIS
Server
Instance
To
continue,
click
N
Cluster
Address timeout:
IPRemove
Ad 160
<Job IP
completion
secondsCluster Name
Add ->
IP Address
Disk G:
Message Queuing
Network Name
<- Remove
NNTP Server Instance
Physical Disk
< Back
Next >
Cancel
< Back
Next >
< Back
Reso
Netw
Phys
Cancel
Next >
< Back
Cancel
Finish
Cancel

You use the print spooler resource type to create a print
spooler for the cluster so that you can use the cluster as
a print server. Without a print spooler resource, it is not
possible to add printers to a cluster. This resource has
two dependencies, a cluster name and a disk
dependency. Multiple print spooler resources can exist
in a cluster, but each group can have no more than one
print spooler resource.

The Print Spooler Parameters property page contains the
following configuration options.
Option
Description
Spool folder
The drive letter and path to where the
spool files will be stored.
Job completion timeout
How long the document can take to
move from the computer to the printer
before the printer stops trying to print
the document.

When a group containing a print spooler resource fails
over to another node, the document that is currently
being spooled to the printer is restarted from the other
node after failover. When you move a print spooler
resource or take it offline, Cluster service waits until all
of the jobs that are spooling to the printers are finished
(or until the configured wait time has elapsed). If the
group containing the print spooler resource fails over
before the application has finished spooling, documents
that are spooling from an application to a print spooler
resource are discarded. These documents must be
respooled (or reprinted) to the print spooler resource.
Print Spooler Entries in the Cluster Registry

Two resource-specific entries are stored under the resource’s
globally unique identifier (GUID) in
HKEY_LOCAL_MACHINE\Cluster\Resources:



JobCompletionTimeout. The amount of time (in milliseconds) that
the print spooler will wait for the currently printing jobs to complete
before the print spooler resource goes offline. This value should be
less than the PendingTimeout for the resource, so that the spooler
has time to delete the pending jobs.
DefaultSpoolDirectory. The full path of the folder in which spool
files are kept. The location of the spool folder will default to a
directory called Spool on the physical disk resource on which the
print spooler resource depends, for example, W:\Spool.
Note: Your spool directory should be different from your data and
application directories to reduce fragmentation on your spool
directory.
Adding a Print Share
\\cluster
File
Edit
View Favorites
Tools
Search
Back
Help
Folders
History
\\cluster
Go
Printers on cluster
cluster
File
Edit
Printers on cluster Back
SYSVOL
NETLOGON
View Favorites Tools Help
Search
Folders
This folder contains information
about printers that are currently
installed, and a wizard to help you
install new printers.
Printers
Scheduled
Tasks
History
Go
Printers on cluster
Comment: Adds, removes, and
configures local and network
printers.
Adds, removes, and configures local and network printers
Printers
Users
Add Printer Wizard
Add Printer
Welcome to the Add Printer
Wizard
This wizard helps you install a printer or make printer
connections.
To continue, click Next.
1 object(s)
< Back
Next >
Cancel

Adding a print share to a virtual server is the same
process as adding a print share to a remote physical
server. Each node also needs to have the appropriate
drivers installed.

After you have installed the drivers on both nodes, browse the
network resources of the virtual server containing the print queue
resource. Double-click the Printers icon and double-click the Add
Printer icon to start the Add Printer Wizard. The wizard will give you
the option to create a port to redirect print jobs to a network print
device. You will need to provide a driver and then a share name for
the printer. Each printer share must be unique in every virtual
server and on every node in the cluster. You will also have the
option to add this printer share to Active Directory.

Note: You cannot use all of the printer ports in a server cluster. You
can use LprMon and Standard Transmission Control
Protocol/Internet Protocol (TCP/IP) Port, both provided with
Windows 2000, in a server cluster. You cannot use local ports in a
cluster because they do not provide failover capability.
Port Configuration Stored in Cluster Registry

It is not necessary to create locally defined printers on
each node in the cluster. The port configuration is
stored in the cluster registry and therefore replicated to
the other node of the cluster.
Lab B: Creating a Print Share
Objectives
After completing this lab, you will be able to:

Add a print driver on each node of the cluster.

Add a print spooler resource to the cluster.

Add a printer to the cluster.

Verify that clustered print share will fail over and print.

Test the failover of a paused printer.
Prerequisites

Before working on this lab, you must be familiar with the
concepts of creating a cluster print share in Module 6,
“Creating Cluster Resources,” of Course 2087A. You
must also have a server cluster that is installed and
running on both nodes.
Lab Setup
To complete this lab, you need the following:

Two computers running Microsoft Windows 2000 Advanced Server,
each with a small computer system interface (SCSI) adapter
installed. Each computer in a student pair must have a different
SCSI ID.

A shared external SCSI disk. Each computer is connected to the
shared disk with SCSI cables. The shared external SCSI disk ID
must be different than either of the SCSI cards of the computers in
the student pair.

Two network adapter cards in each node of the cluster.

A Hewlett-Packard LaserJet 5 printer that is configured on London
and shared as HPLaserJ.

Print Services for Unix installed on the instructor’s computers and
the student computers.

Windows 2000 Setup files.
Scenario

In this exercise, you have been asked to create a
reliable printer share on the cluster with failover
capability.

The following exercises will refer to your computers as
Node A and Node B. Both nodes will install the printer
drivers. Node A will install the spooler resource. Node B
will add the printer. Node A will failover the spooler
resource to Node B. Node B will test the failover of a
paused printer
Exercise 1: Creating a Print Share

In this lab you will configure a print share to be used on
a virtual server. First you will configure a print driver on
each node. Then one student will configure the spool
resource on the cluster. The other student will configure
the port and the share on the virtual server.
To add a print driver on each node of the cluster
Perform this task from Node A and Node B.
1. From the Start menu, point to Settings, and then click
Printers.
2. Click File.
3. Click Server Properties.
4. Click the Drivers tab in the Print Server Properties
dialog box.
5. Click Add.
6. Click Next to start the Add Printer Driver Wizard.
7. From the Manufacturers list, select HP.
8. From the Printers list, select HP LaserJet 5.
9. Click Next.
10. Verify that Intel Windows 2000 drivers is selected, and
then click Next.
11. Click Finish.
12. Close the Print Server Properties dialog box.
13. Close the Printers window.
To add a print spooler resource to the Cluster
Perform this task from Node A.
1. Open Cluster Administrator.
2. Click File.
3. Point to New, and then click Resource.
4. Fill out the following information in the New Resource
dialog box: Name: Print Spooler Description: Print
Spooler Resource type: Print Spooler Group: Cluster
Group
5. Click Next.
6. Verify all of the nodes of your cluster are Possible
owners, and then click Next.
7. Add the following from Available Resources to
Resource dependencies by clicking the name of the
resource, and then click Add ->
Cluster Name:
Disk W:
8. Click Next.
9. Verify that the Spool folder: is W:\Spool, and then click
Finish.
10. Click OK to confirm the added resource.
11. Right-click the Print Spooler resource, and then click
Bring Online.
To add a printer to the cluster
Perform this task from Node B.
1. On the Start menu, click Run.
2. In the Run dialog box, type \\Your_Cluster_Name and
then click OK.
3. Double-click the Printers icon.
4. Double-click the Add Printer icon.
To run the Add Printer Wizard
Perform this task from Node B.
1. Click Next to start the Add Printer Wizard.
2. Click Next to add Remote print server
\\Your_Cluster_Name.
3. Click Create a new port: in the Select the Printer Port
dialog box.
4. Select LPR Port from the Type: drop down list, and then
click Next.
5. Fill out the following in the Add LPR compatible printer
dialog box, and then click OK.
Name or address of server providing lpd: london
Name of printer or print queue on that server:
HPLaserJ
6. From the Manufacturers list, select HP.
7. From the Printers list, select HP LaserJet 5.
8. Click Next.
9. Select Keep existing driver (recommended), then and
click Next.
To configure the share name
Perform this task from Node B.
1. In the Print: box, type HP LaserJet on Cluster_Name and then click
Next.
2. In the Share as: box, type HPLaserJ and then click Next.
3. On the Location and Comment page, click Next.
4. On the Print Test Page page, select Yes, and then click Next.
5. Click Finish.
6. Click OK to close the test page print confirmation.
7. Verify that the print job has printed by double-clicking HP LaserJet
5 in the Printers on Cluster_Name window.
8. The test print job should not be in the Document Name in the HP
LaserJet 5 on Cluster Name on Cluster Name dialog box.
To verify that the clustered print share fails over and
can still print
Perform this task from Node A.
1. From Cluster Administrator, right-click Cluster Group, and then click
Move Group.
2. Wait until all of the resources from the cluster group have moved to
the other node and are online.
3. On the Start menu, click Run and type \\Cluster_Name and then click
OK.
4. Double-click the Printers icon (not the HPLaserJ icon).
5. Double-click HP LaserJet 5 on Cluster Name.
6. Click Printer, and then click Properties.
7. Click Print Test Page.
8. Click OK in the HP LaserJet 5 on Cluster Name on Cluster Name
dialog box.
9. Notice that the test page prints to the print queue and successfully
prints to the remote print device.
To test the failover of a paused printer
Perform this task from Node B.
1. On the Start menu, click run and type \\Cluster_Name and then click
OK.
2. Double-click the Printers icon (not the HPLaserJ icon).
3. Right-click HP LaserJet 5 on Cluster Name and then click Pause
Printing.
4. Right-click HP LaserJet 5 on Cluster Name and then click
Properties.
5. Click Print Test Page.
6. Notice that the test page prints to the print queue but does not print
to the remote print device.
7. Click OK to close the test page message.
8. Click OK to close the Printer Properties page.
9. Double-click HP LaserJet 5 on Cluster Name. Notice that the test
page has been paused on the printer.
To move the spooler resource
Perform this task from Node B.
1. From Cluster Administrator, right-click Cluster Group and then click
Move Group.
2. Wait until all of the resources from the cluster group have moved to
the other node and are online.
3. If the Error Opening Printer message box appears, click OK to close.
4. On the Start menu, click run, type \\Cluster_Name and then click OK.
5. Double-click the Printers icon (not the HPLaserJ icon).
6. Double-click HP LaserJet 5 on Cluster Name.
7. Notice that the test page is still paused on the printer.
8. Click Printer, and then clear Pause Printing.
9. Notice that the printer successfully printed the test page from the
other node.
 Configuring Clustered Applications and Services

Creating Generic Applications and Services

Configuring DHCP and WINS

Microsoft BackOffice Applications

You can install cluster-unaware applications and
services on a server cluster if you configure them as
generic resource types. A cluster-unaware application
that is configured as a generic resource type provides
that application restart and failover capabilities. An
application’s licensing requirement might change if
configured to run on a cluster. Be sure to check vendor
licensing requirements.

You can install cluster-aware applications and services
on a server cluster because they have been written to
meet Cluster service specifications. Cluster-aware
applications and services have custom DLLs providing
more information on the state of resource than generic
resource types provide.

There are three types of server applications and
services that will benefit from running on a server
cluster:

Generic applications and services. Cluster service
includes a wizard for setting up any cluster-unaware
server application for basic error detection, automatic
recovery, and operator-initiated management (such as
moving the application from one node to the other). A
cluster-unaware application that performs best in a
cluster is one that keeps a recoverable state on shared
disk(s), and whose client can handle a pause in service
of up to a minute as Cluster service automatically
restarts the application.

Windows 2000 Advanced Server Services. These
services include, file shares, print queues, Internet and
intranet sites that are managed by Microsoft Internet
Information Services, Microsoft Message Queuing,
Microsoft Component Services, Windows Internet Name
Service (WINS) and Dynamic Host Configuration
Protocol (DHCP).

Cluster-aware applications. Some Microsoft
BackOffice® applications are cluster-aware. There are
also cluster-aware application products available from
other vendors.
Creating Generic Applications and Services
App1 Properties
General
Dependencies
Service1 Properties
Advanced
Parameters Registry Replication
General
App1
Root Registry Key
App1 Properties
Parameters Registry Replication
Programs or services may store data in the registry. Therefore, it is important
to have this data available on the node on which they are running. Specify
the registry keys below HKEY_LOCAL_MACHINE that should be replicated
to all nodes in the cluster.
Root Registry Key
Service1 Properties
Dependencies
Advanced
Parameters Registry Replication
App1
Command line:
Advanced
Service1
Programs or services may store data in the registry. Therefore, it is important
to have this data available on the node on which they are running. Specify
the registry keys below HKEY_LOCAL_MACHINE that should be replicated
to all nodes in the cluster.
General
Dependencies
General
Dependencies
Advanced
Parameters Registry Replication
Service1
w:\app1
Service name:
Current directory: w:\
Service1
Start parameters:
Allow application to interact with desktop
Use Network Name for computer name
Use Network Name for computer name
Add…
Modify…
OK
OK
Cancel
Cancel
Remove
Apply
Apply
Add…
OK
OK
Modify…
Cancel
Cancel
Remove
Apply
Apply

The steps for creating a generic application or service
are similar in that you must name the resource,
configure dependencies, and configure failover policies
for both resource types. Both resource types also have
the ability to replicate local registry information between
the nodes. The configuration of the parameters is
different between the two resource types in that the
application refers to a command line and a directory
path, and the service refers to a service name and
startup parameter.
Parameters for a Generic Application

You use the generic application resource type to
configure cluster-unaware applications for use with
Cluster service. The Parameters property page for the
generic application resource type contains the following
configuration options.
Option
Description
Command line
The command line for the application. The path should
be a location on a shared small computer system
interface drive for failover to take place reliably.
Current directory
The working directory for the application.
Allow application to interact When you select this check box, the application will
with desktop
appear on the desktop of the node on which it is
running.
Use Network Name for
computer name
When you select this check box, the dependent
network name for the group that the service resource
is a member of is used as the computer name.

Note: You must install the service and in the registry of
both nodes before you can configure it as a generic
service.
Potential Application Problems

Applications that you use to launch other applications
may not run correctly in a cluster. It may be necessary
to obtain assistance from the application vendor to
configure the application correctly for the Cluster
service environment.

Cluster service does not support failover of Active
Directory-enabled applications, because you cannot
install Active Directory on the shared disk.
Parameters for a Generic Service

You use the generic service resource type to configure clusterunaware services for use with Cluster service. The Parameters
property page for the generic service resource type contains
the following configuration options.
Option
Description
Service name
The service name as it appears in the Services program in
Control Panel or the Services snap-in, or more technically
the DisplayName that is defined for the service in
HKEY_LOCAL_MACHINE\SYSTEM\
CurrentControlSet\Services.
Startup parameters
Any switches or parameters for the service.
Use Network Name
for computer name
When you select this check box, the network name for the
group that the service resource is a member of is used as
the computer name.

Note: You must register the service before you can configure it.
Registry Replication

The Registry Replication property page is useful for
applications and services that store information in the
HKEY_LOCAL_MACHINE registry tree. This property
page is used to specify the registry keys for the service
that will be replicated to all of the nodes in the cluster.

For example, if an application stores data in
HKEY_LOCAL_MACHINE\Software\MyCompany\MyApp,
you would enter Software\MyCompany\MyApp as the
Root registry key in the dialog box.
Configuring DHCP and WINS

Dynamic Host Configuration Protocol (DHCP)

Windows Internet Name Service (WINS)

As with generic services, you first install WINS or DHCP
on both nodes of the cluster. After you have installed
the DHCP and/or WINS service and the Cluster service,
you need to configure them as a cluster resource type.
The WINS/DHCP resources each have three
dependencies: an IP address, a network name, and a
disk. The WINS or DHCP snap-in can administer these
resources from the virtual server.
To configure a WINS/DHCP resource:
1.
In Cluster Administrator, create a new resource type as either
DHCP Service or WINS Service.
2.
Type a name and description for the resource, and select the group
to which it will belong.
3.
Select the Possible owners.
4.
On the Dependencies tab, add an IP address, network name, and a
disk resource.
5.
On the Parameters tab, select the directory on the shared disk
where the WINS and DHCP database files will be located.
6.
The resource should appear in an offline state in the correct group.
To start the resource, right-click it, and then click Bring Online.
7.
To verify that you have properly configured the resource in the
cluster, right-click the group, and then click Move Group. If the
group moves successfully, you have correctly created and
configured the resource.
Microsoft BackOffice Applications

Microsoft SQL Server Enterprise

Microsoft Exchange Server Enterprise Edition

Microsoft Internet Information Services (IIS)

Message Queuing

You can install some products of the Microsoft
BackOffice suite into a cluster. Other BackOffice
applications can still run on each node of the cluster
and perform work outside of the cluster, but they will
not have failover capability. Microsoft Exchange and
Microsoft SQL Server™ Enterprise Edition, are the
BackOffice applications that are most commonly
installed in a server cluster. They each have different
Cluster service installation procedures.

Depending on which version of SQL Server or Exchange
you are installing on the cluster, you will follow different
installation procedures:

SQL Server creates a virtual server, installs all of the
necessary files on both nodes, and creates all of the
necessary cluster resources running SQL Server.

Exchange 2000 creates additional cluster resources that
need to be manually added if you are going to configure
them to be configured to run on a virtual server.

Other Microsoft BackOffice Applications that you can
install on the server cluster are Microsoft Internet
Information Services (IIS) and Microsoft Message
Queuing.
Using Cluster Application Wizard
Cluster Administrator – [MYCLUSTER (mycluster)]
File View
Window
Help
Open Connection…
Close
Ctrl+O
Bring Online
Take Offline
Initiate Failure
Change Group
1 mycluster
2 server1
3 CLUSTER1
Exit
Active Groups
Active Resources
Network Interfaces
Creates a new cluster group
State
Online
Owner
SERVER2
Resource Type
IP Address
Cluster Name
Online
SERVER2
Network Name
Online
SERVER2
Physical Disk
Online
SERVER2
File Share
Disk W:
Cluster Application Wizard
Public
New
Configure Application
Delete
Rename
Properties
Name
Cluster IP Address
Ctrl+D
Description
Public Share
Welcome to the Cluster
Application
Wizard
Cluster Application Wizard
This wizard helps you to configure an application to run in a
Application
cluster.
It takes youResource
through theType
steps of creating or
Specify
the
typeand
of application
selecting a virtual server,
preparing theresource.
application.
An application is run in a cluster in order to maximize the
The
type identifies
behavior
andon
requirements of a cluster resource. For
availability
of resource
the application
to clientsthe
if the
computer
specify thatitself
this cluster
which it isexample,
running itormay
the application
fails. resource must depend upon another cluster
resource of a specific type.
To continue, click Next.
Resource type: Generic Application
DHCP Service
Distributed Transaction Coordinator
File Share
Generic Application
Generic Service
IIS Server Instance
IP Address
NextQueuing
>
< Back Message
Cancel
Network Name
NNTP Server Instance
Physical Disk
Print Spooler
SMTP Server Instance
< Back
Next >
Cancel

You can manually create a clustered resource, or you
can use the Cluster Application Wizard, which is part of
Cluster service. To launch the wizard, click File on the
menu, and click Configure Application.

If you are creating a new virtual server, you will be
prompted to answer all of the necessary information to
create a virtual server (associate it with a new or
existing group, provide the name, description, and IP
number, and configure failover and failback policies for
the new resources). On the Application Resource Type
page of the wizard, you will select the type of resource
that you are configuring.

If you are using an existing virtual server, you select the
group where you want to locate the resource from the
list of groups that are associated with a virtual server
and then select the type of resource that you are
creating from the list.

After you have selected the type of resource, the wizard
will display the appropriate pages for that resource. You
then complete the creation of the resource, using the
same information that you used when manually creating
a new resource.
Lab C: Creating a WINS Resource
Objectives
After completing this lab, you will be able to:

Prepare WINS to run on the cluster.

Create a WINS resource.

Verify that the cluster is running the WINS service.
Prerequisites

Before working on this lab, you must be familiar with the
concepts of Configuring DHCP and WINS in Course
2087A, Module 6, “Creating Cluster Resources.” You
must also have a server cluster installed and running on
both nodes.
Lab Setup
To complete this lab, you need the following:

Two computers running Microsoft Windows 2000
Advanced Server, each with a small computer system
interface (SCSI) adapter installed. Each computer in a
student pair must have a different SCSI ID.

A shared external SCSI disk. Each computer is
connected to the shared disk with SCSI cables. The
shared external SCSI disk ID must be different than
either of the SCSI cards of the computers in the student
pair.

Two network adapter cards in each node of the cluster.

Windows 2000 Setup files.
Scenario

In this exercise, you have been asked to create a faulttolerant WINS server for your network.

The following exercise will refer to your computers as
Node A and Node B. Node A and Node B will both install
the WINS service. Node A will create the WINS resource
and Node B will verify that the WINS service is running
on the cluster.
Exercise 1: Creating a WINS Resource

In this exercise you will install the WINS service on both
nodes and configure a WINS resource type in the
cluster. You will then verify that the WINS service is
running on the cluster.
To prepare WINS to run on the cluster
Complete this task from Node A and Node B.
1. Click Start, select Settings, and then click Control Panel.
2. In Control Panel, double-click Add/Remove Programs.
3. Click Add/Remove Windows Components.
4. From the Windows Components Wizard, in Components: doubleclick Networking Services.
5. Select the Windows Internet Name Service (WINS) check box, and
then click OK.
6. Click Next.
7. Click Finish.
8. If a Files needed dialog box appears, ask your instructor what
pathname to specify in the Copy files from: field.
9. Close all of the windows.
To create a WINS resource
Complete this task from Node A.
1. Click Start, point to Programs, point to Administrative Tools, and
then click Cluster Administrator.
2. In the console tree, expand Groups.
3. Right-click Cluster Group, select New, and then click Resource.
4. Fill out the following information in the New Resource dialog
box:
Name: WINS
Description: WINS
Resource type: WINS
Service Group: Cluster Group
5. Click Next.
6. Click Next from the Possible Owners dialog box.
7. In the Dependencies dialog box, add the following
Available resources to the Resource Dependencies by
clicking Add:
Cluster IP Address
Cluster Name
Disk W:
8. Click Next.
9. Click Finish from the WINS Service Parameters dialog
box.
10. Click OK in response the confirmation.
11. Right-click Cluster Group, and then click Bring Online.
To verify that your cluster is running the WINS
service
Complete this task from Node B.
1. Click Start, point to Programs, point to Administrative
Tools, and then click WINS.
2. Notice that your cluster is running the WINS service and
can accept WINS name resolution from clients.
3. Close the WINS utility.
Review

Creating a File Share Resource

Creating a Cluster Print Share

Configuring Clustered Applications and Services

Using Cluster Application Wizard