Load Balancing

Load balancer functionality is one of the foundational building blocks of the Citrix NetScaler load balancer. Load balancing is a critical technology to ensure 100 percent application availability and scalability. Any load balancer evaluation must consider the capabilities and sophistication of the load balancer’s core traffic management support, including layer 4 load balancing, layer 7 load balancing (also known as content switching), and global server load balancing (GSLB).

Layer 4 load balancing

The ability to direct incoming client requests to the correct server resource based on layer 2-4 information (e.g., MAC/IP address and TCP port) is mandatory for all load balancers. Load balancers should also include health monitoring, session persistence, and network integration. The NetScaler load balancer delivers the industry’s broadest array of load balancing technologies, including:

Health monitoring

The NetScaler load balancer continuously monitors the availability and health of not only the server hardware, but the state of back-end databases and applications. Network links, operating systems, and even individual application elements are also monitored by the NetScaler load balancer. Some load balancers simply provide basic ping, TCP, and UDP checks, but the NetScaler load balancer goes further to include scriptable health checks, dynamic server response times, and extended content verification (ECV.) The NetScaler load balancer ensures optimal load balancing of sessions to only those servers that are fully functional, preventing bottlenecks and needless session redirects.

Session persistence

The NetScaler load balancer manages persistence requirements for application sessions. This load balancing functionality is required for applications where state information is not automatically shared across a server farm, and each user session must be handled by the same server. Standard load balancers utilize source IP, cookies, and attribute hashing to ensure a sticky connection. The NetScaler load balancer goes further, supporting SSL ID, SIP CALLID, token-based, JSESSIONID, and other advanced routines—all integrated into a single load balancer policy.

Network integration

Most load balancers are generally compatible with the standard routing protocols, such as OSPF, RIP, and BGP, and networking techniques including 802.3ad link aggregation and VLAN tagging. NetScaler load balancers go further, and support dynamic routing, IPv6, and are enhanced with load balancing of SIP and RSTP traffic. NetScaler load balancing techniques include not only rudimentary round robin and least packets, but include server application state protocol (SASP), SNMP-provided metrics, and numerous hashing schemes.

Layer 7 load balancing

Also referred to as content switching, L7 load balancing is essentially an extension of the traffic distribution, health monitoring, and session persistence capabilities of layer 4 load balancers. The advantage is that load balancer decisions can also be based on application layer data. NetScaler load balancers redirect traffic not only on attributes such as HTTP header, URI, SSL session ID, and HTML form data, but can use any payload value. This difference enables the load balancer to improve datacenter efficiency by reducing duplication of server and database resources, leading to improved ROI. With the NetScaler load balancer, all of the services and components that comprise an application no longer need to be implemented on all of the server nodes present. Each physical system can now be tailored to the functions it will be supporting.

Global server load balancing

The NetScaler load balancer takes the layer 4-7 load balancer capabilities and applies them worldwide across geographically distributed server farms. Global server load balancing provides critical business continuity and disaster recovery support in case of site level disruptions and outages.

Such global load balancing further improves performance as remote users have their sessions routed to the closest or best performing datacenter. Resources are now optimized on an enterprise wide basis. Load balancer algorithms include geographic and network proximity to provide the best user response. The NetScaler load balancer continuously monitors numerous site level health attributes to ensure optimum load balancing of global traffic.

If a datacenter goes off line, each user’s session is transparently redirected by the NetScaler load balancer to another location without the user having any knowledge of the changeover. Unlike many other load balancer solutions on the market, the NetScaler load balancer fully integrates global server load balancing into the local load balancer policy. A separate, standalone global server load balancer device is not required.