Product Documentation
ADC
14.1 - Current Release 13.1 13.0 12.1
View PDF

How load balancing works

August 1, 2023
Contributed by: 
S

In a basic load balancing setup, clients send their requests to the IP address of a virtual server configured on the NetScaler appliance. The virtual server distributes them to the load-balanced application servers according to a preset pattern, called the load balancing algorithm. Sometimes, you might want to assign the load balancing virtual server a wildcard address instead of a specific IP address. For instructions about specifying a global HTTP port on the appliance, see Global HTTP Ports.

Load balancing basics

A load balancing setup includes a load-balancing virtual server and multiple load-balanced application servers. The virtual server receives incoming client requests, uses the load balancing algorithm to select an application server, and forwards the requests to the selected application server. The following conceptual drawing illustrates a typical load balancing deployment. Another variation involves assigning a global HTTP port.

Figure 1. Load Balancing Architecture

Load balancing architecture

The load balancing virtual server can use several algorithms (or methods) to determine how to distribute load among the load-balanced servers that it manages. The default load balancing method is the least connection method, in which the NetScaler appliance forwards each incoming client connection to whichever load-balanced application server currently has the fewest active user connections.

The entities that you configure in a typical NetScaler load balancing setup are:

  • Load balancing virtual server. The IP address, port, and protocol combination to which a client sends connection requests for a particular load-balanced website or application. If the application is accessible from the Internet, the virtual server IP (VIP) address is a public IP address. If the application is accessible only from the LAN or WAN, the VIP is usually a private (ICANN non-routable) IP address.
  • Service. The IP address, port, and protocol combination used to route requests to a specific load-balanced application server. A service can be a logical representation of the application server itself, or of an application running on a server that hosts multiple applications. After creating a service, you bind it to a load balancing virtual server.
  • Server object. A virtual entity that enables you to assign a name to a physical server instead of identifying the server by its IP address. If you create a server object, you can specify its name instead of the server’s IP address when you create a service. Otherwise, you must specify the server’s IP address when you create a service, and the IP address becomes the name of the server.
  • Monitor. An entity on the NetScaler appliance that tracks a service and ensures that it is operating correctly. The monitor periodically probes (or performs a health check on) each service to which you assign it. If the service does not respond within the time specified by the time-out, and a specified number of health checks fail, that service is marked DOWN. The NetScaler appliance then skips that service when performing load balancing, until the issues that caused the service to quit responding are fixed.

The virtual server, services, and load balanced application servers in a load balancing setup can use either Internet Protocol version 4 (IPv4) or Internet Protocol version 6 (IPv6) IP addresses. You can mix IPv4 and IPv6 addresses in a single load balancing setup.

For variations in the load balancing setup, see the following use cases:

Understanding the topology

In a load balancing setup, the load balancing server is logically located between the client and the server farm, and manages traffic flow to the servers in the server farm. On the NetScaler appliance, the application servers are represented by virtual entities called services. The following diagram shows the topology of a basic load balancing configuration.

Figure 2. Basic Load Balancing Topology

Load balancing topology

In the diagram, load balancing is used to manage traffic flow to the servers. The virtual server selects the service and assigns it to serve client requests. Consider a scenario where the services Service-HTTP-1 and Service-HTTP-2 are created and bound to the virtual server named Vserver-LB-1. Vserver-LB-1 forwards the client request to either Service-HTTP-1 or Service-HTTP-2. The NetScaler appliance uses the least connection load balancing method to select the service for each request. The following table lists the names and values of the basic entities that must be configured on the appliance.

Entity Name IPAddress Port Protocol
Virtual server Vserver-LB-1 10.102.29.60 80 HTTP
Services Service-HTTP-1 10.102.29.5 80 HTTP
Service-HTTP-2 10.102.29.6 80
Monitors Default None None None

The following diagram shows the load balancing sample values and mandatory parameters that are described in the preceding table.

Figure 3. Load Balancing Entity Model

Load balancing entity model

Use of wildcards instead of IP addresses and ports

Sometimes you might need to use a wildcard for the IP address or the port of a virtual server or for the port of a service. The following cases might require using a wildcard:

  • If the NetScaler appliance is configured as a transparent pass through, which must accept all traffic that is sent to it regardless of the IP or port to which it is sent.
  • If one or more services listen on ports that are not well known.
  • If one or more services, over time, change the ports that they listen on.
  • If you reach the limit for the number of IP addresses and ports that you can configure on a single NetScaler appliance.
  • If you want to create virtual servers that listen for all traffic on a specific virtual LAN.

When a wildcard-configured virtual server or service receives traffic, the NetScaler appliance determines the actual IP address or port and creates records for the service and associated load balanced application server. These dynamically created records are called dynamically learned server and service records.

For example, a firewall load balancing configuration can use wildcards for both the IP address and port. If you bind a wildcard TCP service to this type of load balancing virtual server, the virtual server receives and processes all TCP traffic that does not match any other service or virtual server.

The following table describes some of the different types of wildcard configurations and when each must be used.

IP Port Protocol Description
* * TCP A general wildcard virtual server that accepts traffic sent to any IP address and port on the NetScaler appliance. When using a wildcard virtual server, the appliance dynamically learns the IP and port of each service and creates the necessary records as it processes traffic.
* * TCP A firewall load balancing virtual server. You can bind firewall services to this virtual server, and the NetScaler appliance passes traffic through the firewall to the destination.
IP Address * TCP, UDP, and ANY A virtual server that accepts all traffic that is sent to the specified IP address, regardless of the port. You must explicitly bind to this type of virtual server the services to which it will redirect traffic. It does not learn them dynamically.
* port SSL, SSL_TCP A virtual server that accepts all traffic sent to any IP address on a specific port. Used for global transparent SSL offloading. All SSL, HTTP, and TCP processing that usually is performed for a service of the same protocol type is applied to traffic that is directed to this specific port. The appliance uses the port to dynamically learn the IP of the service it must use. If –cleartext is not specified, the NetScaler appliance uses end-to-end SSL.
* port Not applicable All other virtual servers that can accept traffic to the port. You do not bind services to these virtual servers. The NetScaler appliance learns them dynamically.

Note: If you have configured your NetScaler appliance as a transparent pass through that uses global (wildcard) ports, you might want to turn on Edge mode. For more information, see “Configuring Edge Mode.”

The NetScaler appliance attempts to locate virtual servers and services by first attempting an exact match. If none is found, it continues to search for a match based on wildcards, in the following order:

  1. Specific IP address and specific port number
  2. Specific IP address and a * (wildcard) port
    • (wildcard) IP address and a specific port
    • (wildcard) IP address and a * (wildcard) port

If the appliance is unable to select a virtual server by IP address or port number, it searches for a virtual server based on the protocol used in the request, in the following order:

  1. HTTP
  2. TCP
  3. ANY

Configuring global HTTP ports

You do not configure services or virtual servers for a global HTTP port. Instead, you configure a specific port by using the set ns param command. After configuring this port, the NetScaler appliance accepts all traffic that matches the port number, and processes it as HTTP traffic, dynamically learning and creating services for that traffic.

You can configure more than one port number as a global HTTP port. If you are specifying more than one port number in a single set ns param command, separate the port numbers by a single white space. If one or more ports have already been specified as global HTTP ports, and you want to add one or more ports without removing the ports that are currently configured, you must specify all the port numbers, current and new, in the command. Before you add port numbers, use the show ns param command to view the ports that are currently configured.

To configure a global HTTP port by using the command line interface

At the command prompt, type the following commands to configure a global HTTP port and verify the configuration:

set ns param –httpPort <port>

show ns param
<!--NeedCopy-->

Example 1: Configuring a port as a global HTTP port

In this example, port 80 is configured as a global HTTP port.

set ns param -httpPort 80
Done
show ns param
         Global configuration settings:
                          HTTP port(s): 80
                       Max connections: 0
           Max requests per connection: 0
                   Client IP insertion: DISABLED
                        Cookie version: 0
        Persistence Cookie Secure Flag: ENABLED
        ...
        ...
<!--NeedCopy-->

Example 2: Adding ports when one or more global HTTP ports are already configured**

In this example, port 8888 is added to the global HTTP port list. Port 80 is already configured as a global HTTP port.

> show ns param
        Global configuration settings:
                          HTTP port(s): 80
                       Max connections: 0
           Max requests per connection: 0
                   Client IP insertion: DISABLED
                        Cookie version: 0
        Persistence Cookie Secure Flag: ENABLED
                          Min Path MTU: 576
        ...
        ...
 Done
> set ns param -httpPort 80 8888
 Done
>  show ns param

        Global configuration settings:
                          HTTP port(s): 80,8888
                       Max connections: 0
           Max requests per connection: 0
                   Client IP insertion: DISABLED
                        Cookie version: 0
        Persistence Cookie Secure Flag: ENABLED
                          Min Path MTU: 576

        ...
        ...
 Done
>
<!--NeedCopy-->

To configure a global HTTP port by using the configuration utility

  1. Navigate to System > Settings > Change HTTP Parameters, and then add an HTTP port number.
How load balancing works
August 1, 2023
Contributed by: 
S
August 1, 2023
Contributed by: 
S

In this article

Site feedback | Your privacy choices footer linkYour Privacy Choices | Privacy and legal terms | Cookie preferences | Consent settings | docs.cloud.com
© 1999- Cloud Software Group, Inc. All rights reserved.