In this article I will
load balance 2 servers and take you through the process step-by-step. Load
Balancing takes 2 or more servers and lets them share one IP address so both
servers can serve client requests. At the end of this article you should be
able to configure NLB.
Gathering Information
Log onto both of the
servers and run IPCONFIG /ALL from the command prompt. We need the name, domain
and IP address of each server that will be in the NLB Cluster. We will also
need to make up an additional name for the cluster in this example we will use
SERVER-LB for the virtual cluster name.
The 2 servers we will be Load Balancing are PL2008-01 and
PL2008-02. The virtual cluster name will be PL2008-V. So if this was a web
server users would go to http://PL2008-V, depending how we configure NLB either
PL2008-01, PL2008-02 or both servers will service the web request.
SERVER
NAME
|
IP
ADDRESS
|
TYPE
|
PL2008-01.pintolake.net
|
192.168.1.180
|
Server 1
|
PL2008-02.pintolake.net
|
192.168.1.181
|
Server 2
|
PL2008-V.pintolake.net
|
192.168.1.182
|
Virtual cluster name and IP
address of Servers 1/2
|
In this example both
servers only have one network card. If you have multiple network cards you will
still be able to load balance the 2 servers. You need to configure one NIC per
server for NLB, both NIC’s should be on the same VLAN and be they should able
to contact each other.
PL2008-01
PL2008-02
Installation of NLB
feature on all NLB nodes
This should be done on
ALL NODES in the NLB Cluster. In this case we are performing this installation
on PL2008-01 and PL2008-02.
Open Server Manager,
you can open this several different ways in Windows Server 2008. Probably the
quickest way to open Server Manager is to right click "My Computer"
and choose "Manage", another way is open "Control Panel" go
to "Program and Features" and select "Turn Windows features on
or off". A third way to open it is "Server Manager" option under
Administrative Tools.
·
Select
"Features" from the Server Manager menu on the left
·
Press "Add
Features"
·
Select the checkbox
next to "Network Load Balancing"
·
Press "Next"
·
Press
"Install"
Installation will
proceed to install the necessary components
Installation has
successes. It is highly recommended that you repeat this process on all nodes
in the NLB cluster at this point before continuing with configuration
·
Press
"Close"
NOTE: Network Load Balancing may also be installed from a
command prompt with elevated privileges (right click on the command prompt in
the Start menu and select Run as administrator) by running theservermanagercmd
-install nlb command.
For example:
C:\Windows\system32>servermanagercmd -install nlb
......
Start Installation...
[Installation] Succeeded: [Network Load Balancing].
<100>
Success: Installation succeeded.
|
Configuring NLB on
NODE 1 (PL2008-01)
Network Load Balanced
clusters are built using the Network Load Balancing Manager which
you can start from Start -> All Programs -> Administrative Tools menu
or from a command prompt by executing nlbmgr.
·
Under the Cluster Menu
option select "New"
·
Enter the first node
in the cluster which is PL2008-01
·
Press
"Connect"
You will have the
option to choose which network adapter you want to use, the NIC should be on the
same subnet as the other servers in the NLB cluster
·
Press "Next"
·
Enter the Priority ID
as, 1 (each node in the NLB cluster should have a UNIQUE ID)
·
Make sure the correct
adapter was selected under "Dedicated IP Address"
·
Select
"Started" for the "Initial host state" (this tells NLB
whether you want this node to participate in the cluster at startup)
·
Press "Next"
·
Press "Add"
·
Enter the Cluster IP
and Subnet mask
·
Press "OK"
You can add multiple
IP Addresses for the cluster, enter as many as you want.
·
Make sure the
"Cluster IP addresses" are correct
·
Press "Next"
·
Select the IP Address
for this cluster
·
Enter the NLB address
"PL2008-V.pintolake.net"
·
Enter
"Unicast" as the "Cluster operation mode"
·
Press "Next"
Unicast vs Multicast
Unicast/Multicast is the way the
MAC address for the Virtual IP is presented to the routers. In my experience
I have almost always used Multicast, which if you use you should
enter a persistent ARP entry on all upstream switches or you will not be able
to ping the servers remotely.
In the unicast method:
·
The
cluster adapters for all cluster hosts are assigned the same unicast MAC
address.
·
The
outgoing MAC address for each packet is modified, based on the cluster host’s
priority setting, to prevent upstream switches from discovering that all
cluster hosts have the same MAC address.
In the multicast method:
·
The
cluster adapter for each cluster host retains the original hardware unicast
MAC address (as specified by the hardware manufacture of the network
adapter).
·
The
cluster adapters for all cluster hosts are assigned a multicast MAC address.
·
The
multicast MAC is derived from the cluster’s IP address.
·
Communication
between cluster hosts is not affected, because each cluster host retains a
unique MAC address.
Selecting the Unicast or Multicast
Method of Distributing Incoming Requestshttp://technet.microsoft.com/en-us/library/cc782694.aspx
|
I am leaving all the
default for the port rules; by default its set to all ports with Single
affinity, which is sticky. For more information on Port Rules, see my Note
below.
·
Press
"Finish"
NOTE: Add/Edit Port Rule Settings
For most scenarios I would keep
the default settings. The most important setting is probably the filtering
mode. "Single" works well for most web application, it maintains a
users session on one server so if the user server requests go to PL2008-01,
PL2008-02 will continue to serve that request for the duration of the
session.
None
·
You want
to ensure even load balancing among cluster hosts
·
Client
traffic is stateless (for example, HTTP traffic).
Single
·
You want
to ensure that requests from a specific client (IP address) are sent to the
same cluster host.
·
Client
state is maintained across TCP connections (for example, HTTPS traffic).
Class C
·
Client
requests from a Class C IP address range (instead of a single IP address) are
sent to the same cluster host.
·
Clients
use multiple proxy servers to access the cluster, and they appear to have
multiple IP addresses within the same Class C IP address range.
·
Client
state is maintained across TCP connections (for example, HTTPS traffic).
For more information on this please
see this TechNet article:
Specifying the Affinity and
Load-Balancing Behavior of the Custom Port Rule
http://technet.microsoft.com/en-us/library/cc759039.aspx
 |
You should see a
couple of things in the NLB Manager, this will let us know that this node successfully
converged on our new PL2008-V.pintolake.net NLB Cluster
·
Make sure the node’s
status changes to "Converged"
·
Make sure you see a
"succeeded" message in the log window
Configuring NLB for
NODE 2 (PL2008-02)
We will configure
PL2008-02 from PL2008-01. If we wanted to configure this from PL2008-02 then we
would need to connect to the PL2008-V cluster first then add the host to the
cluster.
·
Right click the
cluster name "PL2008-V.pintolake.net" and select "Add Host to
Cluster"
·
Enter PL2008-02 and press
"Connect"
A list of Network
adapters will show up
·
Select the network
adapter you want to use for Load Balancing
·
Press "Next"
This step is very
important; each node in the NLB cluster should have a unique identifier. This
identifier is used to identify the node in the cluster.
·
Enter the Priority ID
as, 2 (each node in the NLB cluster should have a UNIQUE ID)
·
Make sure the correct
adapter was selected under "Dedicated IP Address"
·
Select
"Started" for the "Initial host state" (this tells NLB
whether you want this node to participate in the cluster at startup)
·
Press "Next"
·
Press
"Finish"
You should see a
couple of things in the NLB Manager, this will let us know that both nodes
successfully converged on our new PL2008-V.pintolake.net NLB Cluster
·
Make sure that both
node’s status changes to "Converged"
·
Make sure each node
has a unique "host priority" ID
·
Make sure each node is
"started" under "initial host state"
·
Make sure you see a
"succeeded" message in the log window for the second node
A closer look at the
configuration information for this NLB cluster
Testing
·
Go to the command
prompt and type "wlbs query", as you can see HOST 1 and HOST 2
converged successfully on the cluster. This means things are working well.
·
Ping each server
locally and remotely
·
Ping the virtual IP
locally and remotely – you should do this three times from each location. If
you cannot ping remotely you may need to add a static ARP entry in your
switches and/or routers where the host machines reside
o 1 – Both nodes up
o 2 – Node 1 down
o 3 – Node 2 down
NLB Documentation
(from Windows Help)
Availability,
scalability, and clustering technologies
Windows Server 2008
provides two clustering technologies: failover clusters and Network Load
Balancing (NLB). Failover clusters primarily provide high availability; Network
Load Balancing provides scalability and at the same time helps increase
availability of Web-based services.
Your choice of cluster
technologies (failover clusters or Network Load Balancing) depends primarily on
whether the applications you run have long-running in-memory state:
Failover clusters are
designed for applications that have long-running in-memory state, or that have
large, frequently updated data states. These are called stateful applications,
and they include database applications and messaging applications. Typical uses
for failover clusters include file servers, print servers, database servers,
and messaging servers.
Network Load Balancing
is intended for applications that do not have long-running in-memory state.
These are called stateless applications. A stateless application treats each
client request as an independent operation, and therefore it can load-balance
each request independently. Stateless applications often have read-only data or
data that changes infrequently. Front-end Web servers, virtual private networks
(VPNs), File Transfer Protocol (FTP) servers, and firewall and proxy servers
typically use Network Load Balancing. Network Load Balancing clusters can also
support other TCP- or UDP-based services and applications.
Network Load Balancing
overview
The Network Load
Balancing (NLB) service enhances the availability and scalability of Internet
server applications such as those used on Web, FTP, firewall, proxy, virtual
private network (VPN), and other mission-critical servers.
What are NLB clusters?
A single computer
running Windows can provide a limited level of server reliability and scalable
performance. However, by combining the resources of two or more computers
running one of the products in Windows Server 2008 into a single virtual
cluster, NLB can deliver the reliability and performance that Web servers and
other mission-critical servers need.
Each host runs a
separate copy of the desired server applications (such as applications for Web,
FTP, and Telnet servers). NLB distributes incoming client requests across the
hosts in the cluster. The load weight to be handled by each host can be
configured as necessary. You can also add hosts dynamically to the cluster to
handle increased load. In addition, NLB can direct all traffic to a designated
single host, which is called the default host.
NLB allows all of the
computers in the cluster to be addressed by the same set of cluster IP
addresses, and it maintains a set of unique, dedicated IP addresses for each
host. For load-balanced applications, when a host fails or goes offline, the
load is automatically redistributed among the computers that are still
operating. When a computer fails or goes offline unexpectedly, active
connections to the failed or offline server are lost. However, if you bring a
host down intentionally, you can use the drainstop command to service all
active connections prior to bringing the computer offline. In any case, when it
is ready, the offline computer can transparently rejoin the cluster and regain
its share of the workload, which allows the other computers in the cluster to
handle less traffic.
Hardware and software
considerations for NLB clusters
·
NLB is installed as a
standard Windows networking driver component.
·
NLB requires no
hardware changes to enable and run.
·
NLB Manager enables
you to create new NLB clusters and to configure and manage clusters and all of
the cluster’s hosts from a single remote or local computer.
·
NLB lets clients
access the cluster by using a single, logical Internet name and virtual IP
address—known as the cluster IP address (it retains individual names for each
computer). NLB allows multiple virtual IP addresses for multihomed servers.
Note:
In the case of virtual
clusters, the servers do not need to be multihomed to have multiple virtual IP
addresses.
NLB can be bound to
multiple network adapters, which allows you to configure multiple independent
clusters on each host. Support for multiple network adapters is different from
virtual clusters in that virtual clusters allow you to configure multiple
clusters on a single network adapter.
Installing the NLB
feature
To use Network Load
Balancing (NLB), a computer must have only TCP/IP on the adapter on which NLB
is installed. Do not add any other protocols (for example, IPX) to this
adapter. NLB can load balance any application or service that uses TCP/IP as
its network protocol and is associated with a specific Transmission Control
Protocol (TCP) or User Datagram Protocol (UDP) port.
To install and configure
NLB, you must use an account that is listed in the Administrators group on each
host. If you are not using an account in the Administrators group as you
install and configure each host, you will be prompted to provide the logon
credentials for such an account. To set up an account that NLB Manager will use
by default: in NLB Manager, expand the Options menu, and then click
Credentials. We recommend that this account not be used for any other purpose.
You can use Initial
Configuration Tasks or Server Manager to install NLB. To install NLB, in the
list of tasks, click Add features and in the list of features in the wizard,
click Network Load Balancing.
Managing NLB
Server roles and
features are managed by using Microsoft Management Console (MMC) snap-ins. To
open the Network Load Balancing Manager snap-in, click Start, click
Administrative Tools, and then click Network Load Balancing Manager. You can
also open Network Load Balancing Manager by typing Nlbmgr at a command prompt.
Additional references
for NLB
To learn more about
NLB, you can view the Help on your server. To do this, open Network Load
Balancing Manager as described in the previous section and press F1.
For instructions on
how to configure NLB with Terminal Services, see http://go.microsoft.com/fwlink/?LinkId=80406
No comments:
Post a Comment