Apply Your Knowledge


Exercises

5.1 Creating a New NLB Cluster

In this exercise, you create a new network load balancing cluster. You will need a Windows Server 2003 Enterprise computer to complete this exercise.

Estimated time: 20 minutes

  1. Open the Network Connections window.

  2. Configure the IP address and subnet mask for the load balancing and administrative network adapters on the first NLB cluster node.

  3. Open the Network Load Balancing Manager.

  4. Right-click on Network Load Balancing Clusters and select New Cluster from the context menu.

  5. Enter the cluster's IP address, subnet mask, and cluster domain name . Configure the cluster for unicast or multicast as desired. You can also configure IGMP multicast and remote control as desired. Click Next to continue.

  6. On the Cluster IP Addresses dialog box, enter any additional virtual IP addresses using the Add button.

  7. On the Port Rules dialog box, configure any port rules that are appropriate for your NLB cluster installation.

  8. Configure your port rules as required.

  9. On the Connect dialog box, type the name of the first cluster node and click the Connect button. Select the network adapter that is to be used for load balancing.

  10. On the Host Parameters dialog box, configure the host priority, cluster node dedicated IP address and subnet mask, and the initial state of the cluster node.

5.2 Adding Nodes to an Existing NLB Cluster

In this exercise, you add an additional node to an existing NLB cluster. You need to complete Exercise 5.1 before attempting this one.

Estimated time: 15 minutes

  1. On the server that is to be added to the NLB cluster, open the Network Connections.

  2. Open the network adapter Status dialog box for the network adapter that will be used for load balancing.

  3. Click the Properties button to open the network adapter Properties dialog box.

  4. Install the Network Load Balancing service.

  5. From the cluster node where the NLB cluster was created, open the Network Load Balancing Manager.

  6. If the Network Load Balancing Manager does not display the cluster, connect to it.

  7. Right-click on the NLB cluster and select Add Host To Cluster from the context menu.

  8. On the Connect dialog box, type the name of the additional cluster node and click the Connect button. Select the network adapter that is to be used for load balancing.

  9. On the Host Parameters dialog box, configure the host priority, cluster node dedicated IP address and subnet mask, and the initial state of the cluster node.

5.3 Creating an MSCS Cluster

In this exercise, you create a new MSCS cluster. You need a Windows Server 2003 Enterprise computer that is part of an Active Directory domain to accomplish this exercise.

Estimated time: 20 minutes

  1. Create a domain user account to be used by the MSCS service. Configure the password to never expire.

  2. Ensure that the load balancing and administrative network adapters on the first cluster node are configured correctly.

  3. Open the Cluster Administrator. You should be prompted with the Open Connection to Cluster dialog box. If not, click File, Open Connection. Select Create New Cluster and click OK.

  4. Click Next to dismiss the opening dialog box of the New Cluster Creation Wizard.

  5. On the Cluster Name and Domain dialog box, select the cluster domain from the drop-down list. Enter the cluster name in the space provided.

  6. On the Select Computer dialog box, select the computer that will be the first node in the new cluster.

  7. The Analyzing Configuration dialog box appears and runs for a short period of time.

  8. On the IP Address dialog box, enter the IP address that is being assigned to the cluster.

  9. On the Cluster Service Account dialog box, enter the proper credentials for the cluster domain user account you created previously.

  10. On the Proposed Cluster Configuration dialog box, review the cluster configuration before continuing. Change the quorum type if desired.

5.4 Adding a Node to an Existing MSCS Cluster

In this exercise, you add an additional node to an existing MSCS cluster. Before starting this exercise, you need to complete Exercise 5.3 and have a second computer that is part of the same domain as in Exercise 5.3.

Estimated time: 15 minutes

  1. Ensure that the load balancing and administrative network adapters on the new cluster node are configured correctly.

  2. Open the Cluster Administrator. If the cluster does not appear in the Cluster Administrator, click File, Open Connection and supply the required information to connect to the cluster.

  3. Right-click on the cluster name in the Cluster Administrator and select New, Node from the context menu.

  4. Click Next to dismiss the opening dialog box of the Add Nodes Wizard.

  5. On the Select Computers dialog box, enter the computer names that are to be joined to the cluster.

  6. The Analyzing Configuration dialog box appears for the new node(s) providing information about their suitability to join the cluster.

  7. On the Cluster Service Account dialog box, enter the correct password for the cluster service account.

  8. On the Proposed Cluster Configuration dialog box, you can review the cluster configuration before continuing.

  9. The Adding Nodes to the Cluster dialog box appears, detailing the status of the node addition.

  10. Click Finish to complete the Add Nodes Wizard.

Review Questions

1:

What is network load balancing?

2:

What does convergence mean?

3:

What desirable function do port rules provide?

4:

What ports must be made available for NLB remote control to work?

5:

What is the difference between load balancing and clustering?


Exam Questions

1:

Andrea is your newly hired assistant network administrator. She has been asking you recently about high availability solutions using Windows Server 2003. She tells you that she has heard of the phrase "five nines" before but doesn't understand what it means. What is the correct answer that you should tell her in regards to the meaning of the phrase "five nines"?

  1. Five nines means that servers must be up for at least 99,999 minutes per year.

  2. Five nines means that servers cannot ever have 99.999% CPU utilization.

  3. Five nines means that servers must be up for at least 99.999% of the time per year.

  4. Five nines means that all servers must have a minimum of five 90GB hard drives .

2:

Austin is interviewing for the position of assistant server reliability administrator in your organization. You are asking Austin questions during the technical proficiency portion of his interview. You have asked Austin what types of high availability solutions Windows Server 2003 provides. What two correct answers should Austin provide you?

  1. Network load balancing

  2. Windows File Protection

  3. Clustering

  4. Redundant power supplies

3:

When discussing the process of convergence, what are you referring to?

  1. A network communication sent among individual cluster nodes at intervals of no more than 500 milliseconds that is used to determine the status of all cluster nodes.

  2. The process of a cluster group moving back to the preferred node after the preferred node has resumed cluster membership.

  3. The process by which NLB clustering hosts determine a new, stable state among themselves and elect a new default host after the failure of one or more cluster nodes.

  4. The process of a cluster group moving from the currently active node to another still functioning node in the cluster group.

4:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. You are using the unicast cluster operation mode. In what two ways can you prevent switch port flooding from occurring on the NLB cluster?

  1. Configure IGMP multicast support for the NLB cluster.

  2. Place the NLB cluster nodes on a dedicated switch.

  3. Place the NLB cluster nodes on a dedicate VLAN.

  4. Configure single affinity for the NLB cluster.

5:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. You are trying to decide which clustering mode to use for your back-end servers. What two clustering modes exist?

  1. Active/Passive

  2. Passive/Active

  3. Passive/Passive

  4. Active/Active

6:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. Which of the following cluster models is not available in Windows Server 2003?

  1. Single node cluster

  2. Multiple quorum cluster

  3. Single quorum cluster

  4. Majority node set cluster

7:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. Your e-commerce application makes heavy use of cookies to maintain user session state information during the active connection. Which type of filtering mode should you configure to ensure that clients receive the best possible overall service during their connection period?

  1. Single Host

  2. Disable Port Range

  3. Multiple Host

  4. Allow Port Range

8:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. Your e-commerce application makes heavy use of cookies to maintain user session state information during the active connection. Which type of affinity should you configure to ensure that clients receive the best possible overall service during their connection period?

  1. None

  2. Class C

  3. Multiple

  4. Single

9:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. You are using the unicast cluster operation mode. Which of the following statements are true about the load balancing adapters installed in your network load balancing servers? (Choose two correct answers.)

  1. All cluster nodes share an identical unicast MAC address.

  2. All cluster nodes retain their original MAC address. In addition, the load balancing adapters are also assigned an additional multicast MAC address.

  3. If a network load balancing server is using the unicast operational mode and has only one network adapter installed, it will be unable to communicate with the other nodes in the network load balancing cluster.

  4. Inbound client requests are sent to all cluster nodes using the multicast MAC address.

10:

You are preparing to implement a network load balancing solution for your organization's e-commerce Web site. You will be load balancing six Windows Server 2003 computers running IIS on the front end with four Windows Server 2003 computers in a cluster on the back end providing access to an SQL Server 2000 database. You are using the unicast cluster operation mode. How many total IP addresses will you need to assign for the network load balancing portion of the solution?

  1. 1

  2. 6

  3. 7

  4. 12

11:

Chris is creating a new cluster. She has all the required information written down so that she can enter it during the creation process except for the user account that is to be used. What user account should Chris use during the cluster creation process as the cluster service account?

  1. The built-in local Administrator account

  2. An account that is a member of the Domain Admins group

  3. An account that is a member of the Enterprise Admins group

  4. A user account that is configured for the password to never expire

12:

Jeff wants to use the nlb.exe command to manage his NLB cluster from his desktop computer. What must Jeff do before he will able to manage the NLB cluster from his desktop computer?

  1. Ensure that he is a member of the Enterprise Admins group.

  2. Enable remote control on the NLB cluster.

  3. Disable remote control on the NLB cluster.

  4. Configure the required port rules on the NLB cluster.

13:

Dave is a network administrator in a large Windows Server 2003 Active Directory network. He is responsible for managing and maintaining his company's server clusters. On Monday morning during his normal monitoring cycle, he notices that one of the cluster nodes in his Exchange Server cluster has stopped responding to client requests due to the Cluster Service not running. What is the first step that Dave should perform to begin the recovery process on this failed node?

  1. Initiate an ASR restoration.

  2. Evict the node from the cluster.

  3. Initiate a Windows Backup restoration.

  4. Add a new node to the cluster.

14:

Rick has just completed the planning and configuration of a new NLB cluster that is to be used as the front end of an Internet Web site. All the cluster nodes are connected to the same switch on the core network. After several days Rick has noticed an exceptionally high amount of traffic on the switch that he cannot explain. What might be the cause of this excessive traffic?

  1. The switch is unmanaged and therefore cannot map the MAC address to IP address correctly for the servers attached to it.

  2. The switch is experiencing port flooding.

  3. The switch has an incorrectly configured VLAN on it.

  4. The cluster node network adapters are configured for half-duplex operation.

15:

Rick has just completed the planning and configuration of a new NLB cluster that is to be used as the front end of an Internet Web site. All the cluster nodes are connected to the same switch on the core network. After several days Rick has received several complaints from users stating that their shopping carts sometimes lose the contents as they move around the Web site. What is the most likely reason for this problem?

  1. The client affinity is misconfigured.

  2. The port rules are misconfigured.

  3. The operational mode is misconfigured.

  4. The host priorities are misconfigured.


Answers to Review Questions

A1:

Network load balancing distributes all incoming connection requests using a mathematical algorithm to members of the NLB cluster. For more information, see the section "High Availability Solutions."

A2:

Convergence is the process by which NLB clustering hosts determine a new, stable state among themselves and elect a new default host after the failure of one or more cluster nodes. For more information, see the section "High Availability Terminology."

A3:

Port rules can be used to allow traffic only on specific ports to be load-balanced. If you configure explicit allow rules for specific ports and an explicit deny rule dropping all other traffic, only the desired traffic will be load-balanced . For more information, see the section "Port Rules."

A4:

The default UDP control ports are 1717 and 2504 at the cluster VIP. For more information, see the section "Monitoring Network Load Balancing."

A5:

Network load balancing distributes all incoming connection requests using a mathematical algorithm to members of the NLB cluster. Clustering allows several independent servers to act and appear as a single entity to clients. For more information, see the section "High Availability Solutions."


Answers to Exam Questions

A1:

C. The phrase "five nines" refers to server (or service/application) availability and means that the server must be available for client requests at least 99.999% of the time; this equates to only 5.25 minutes of downtime per year. For more information, see the section "Introduction."

A2:

A, C. Windows Server 2003 provides Network Load Balancing and MSCS clustering services to create high availability server solutions. For more information, see the section "High Availability Solutions."

A3:

C. Convergence is the process by which NLB clustering hosts determine a new, stable state among themselves and elect a new default host after the failure of one or more cluster nodes. For more information, see the section "High Availability Terminology."

A4:

B, C. If the multicast operation mode is configured, you can use the new IGMP multicast support function. Because unicast is being used in this case, you have the option of placing the NLB cluster nodes on a dedicated switch or VLAN to eliminate switch port flooding. For more information, see the section "NLB Cluster Operation Mode."

A5:

A, D. You can select from either Active/Passive or Active/Active clustering mode. Active/Passive has one node in the cluster online and providing services. All other nodes are online and in a standby state, ready to assume the services if the active node fails. Active/Active has an instance of the clustered service running on all nodes in the cluster. If a node fails, the service instance is transferred to one of the remaining operational nodes. For more information, see the section "Planning and Implementing MSCS Clusters."

A6:

B. The single mode cluster, single quorum cluster, and majority node set cluster models are supported in Windows Server 2003. There is no multiple quorum cluster model. For more information, see the section "Cluster Models."

A7:

C. By using the multiple host filtering mode, you will allow all cluster nodes to handle client traffic. When the multiple host filtering mode is configured, you can also configure a client affinity setting. Client affinity determines how cluster nodes will handle repeated client requests during the same session. For more information, see the section "Filtering Mode and Affinity."

A8:

D. By configuring the single affinity setting, you will maintain all client requests on the same node for the duration of the session. Single affinity provides the best support for the maintenance of user state data and is often used when applications that are running on the cluster generate cookies. For more information, see the section "Filtering Mode and Affinity."

A9:

A, C. Unicast mode should be used only when the network load balancing servers have two network adapters installed. If only one network adapter is installed, the cluster nodes will not be able to communicate with each other. When the unicast mode is used, all cluster nodes are assigned an identical unicast MAC address that overwrites the MAC address of the load balancing adapter. For more information, see the section "NLB Cluster Operation Mode."

A10:

C. You will need to assign a total of seven IP addresses for the network load balancing portion of this solution. Each load balancing network adapter in the front-end IIS servers requires a dedicated IP address, such as 10.0.25.1110.0.25.16. Also, the entire network load balancing cluster is assigned a virtual IP (VIP) address, such as 10.0.25.10. For more information, see the section "Creating an NLB Cluster."

A11:

D. Chris should use an ordinary domain user account that has been configured to never have its password expire. During the cluster creation process, this user account will be given Local Administrator privileges on all cluster nodes and will also be delegated cluster- related user rights in the domain, including the Add Computer Accounts to the Domain user right. For more information, see the section "Creating a Cluster."

A12:

B. Before the nlb.exe command can be used to remotely manage an NLB cluster over a LAN or WAN, remote control must be enabled on the NLB cluster. For more information, see the section "Monitoring Network Load Balancing."

A13:

B. Dave will need to first evict the failed node from the cluster before he can begin any recovery operation on the node. For more information, see the section "Recovering from Failed Cluster Nodes."

A14:

B. The most likely reason for the excessive amount of traffic on the switch is due to switch port flooding. The mathematical algorithm used by network load balancing sends inbound traffic to every host in the NLB cluster. It does so by preventing the switch that the NLB cluster nodes are attached to from ever associating the NLB cluster's MAC address with a specific port on the switch. This, however, leads to the unwanted side effect of switch port flooding where the switch floods all its ports with all packets inbound to the NLB cluster. For more information, see the section "NLB Cluster Operation Mode."

A15:

A. Rick's problem is most likely that the client affinity is misconfigured. The affinity should be configured for single. This affinity type maintains all client requests on the same node for the duration of the session (connection). This type of affinity provides the best support for the maintaining of user state data and is often used when applications are running on the cluster that generates cookies. For more information, see the section "Filtering Mode and Affinity."


Suggested Readings and Resources

1. Windows Server 2003 Deployment Kit: Planning Server Deployments, http://go.microsoft.com/fwlink/?LinkId=15309.

2. Clustering Services Technology Center, http://www.microsoft.com/windowsserver2003/technologies/clustering/default.mspx.



MCSE Windows Server 2003 Network Infrastructure (Exam 70-293)
MCSE 70-293 Exam Prep: Planning and Maintaining a Microsoft Windows Server 2003 Network Infrastructure (2nd Edition)
ISBN: 0789736500
EAN: 2147483647
Year: 2003
Pages: 151
Authors: Will Schmied

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net