NetScaler load balancing ensures that applications and services are 100 percent available to all users. Application performance is maximized by a set of powerful acceleration capabilities, including intelligent data compression, static and dynamic content caching, and multiple TCP optimizations that improve the efficiency of the network.

Intelligent load balancing

NetScaler load balancing and content switching capabilities improve the efficient use of server resources by intelligently distributing application requests among multiple servers. NetScaler increases server efficiency and enhances application availability by directing each request to the right server.

Advanced L4-7 load balancing

The NetScaler load balancer delivers the industry's broadest array of load balancing technologies, including health monitoring, session persistence, network integration and content switching.
Continue reading about advanced load balancing

Global load balancing

NetScaler manages application traffic among multiple global datacenters, ensuring that geographically distributed users have their sessions routed to the closest or best performing datacenter.
Continue reading about global load balancing

SQL database load balancing

NetScaler DataStream technology intelligently distributes database queries for Microsoft SQL Server and Oracle MySQL infrastructures.
Continue reading about SQL database load balancing

NetScaler AppCache

Citrix NetScaler includes AppCache, a powerful, fully integrated caching solution that stores static and dynamically-generated application content, such as HTML pages, images, reports and web objects, directly in NetScaler memory. Frequently requested content is directly delivered to clients, offloading busy web servers and reducing the number of round-trips to substantially improve the performance of web applications.

AppCache includes the following key features:

• Advanced caching policy controls: Granular, expression-based policy controls to cache targeted static and dynamic content.
• Powerful content expiry mechanisms: Multiple methods to expire content held in NetScaler cache to ensure application users always receive up-to-date content.
• High-capacity memory caching: NetScaler nCore technology offers the industry's highest memory cache capacity—up to 48GB of storage, 12 times more than the caching memory provided by most other application delivery controllers (ADCs).
• External cache redirection: Automatically redirects requests for specific content to external caching systems in closer proximity to the user to further offload back-end servers, speed application responsiveness and reduce expensive bandwidth consumption.

NetScaler AppCompress

Citrix NetScaler accelerates web application performance through advanced HTTP compression technologies, which are supported by all commercial web browsers. NetScaler AppCompress reduces the size of HTTP and HTTPS responses sent from servers to client browsers – before they are transmitted. By sending less data to application clients, NetScaler AppCompress drives significantly faster response times for most web applications and reduces network bandwidth requirements.

NetScaler AppCompress can compress most types of static and dynamic content including those based on HTML and XML as well as MIME types such as text/rtf, and text/CSS, plus application data such as documents used by common productivity tools (e.g. Microsoft Office).

Both GZIP and DEFLATE compression algorithms are supported, and compatibility with the client browser is automatically checked to make sure the browser supports the algorithm for a transparent user experience. All AppCompress configuration takes place in the single NetScaler policy manager for easy deployment.

TCP Optimization

NetScaler provides a host of advanced technologies to optimize TCP protocol behaviors to improve the efficiency of the underlying network infrastructure, while still preserving 100 percent interoperability with all clients.

NetScaler TCP optimizations include:

Limited transmit

The limited transmit algorithm helps avoid expensive timeouts at the sender by allowing the sender to send additional packets after receiving just two duplicate acknowledgements. This is helpful in situations where the fast retransmit algorithm cannot be applied because the sender does not receive three duplicate acknowledgements, a common occurrence when the congestion window of the sender is small (e.g., a timeout will result when a sender's congestion window is three segments and one of the packet gets lost) or if a large number of packets are lost in a window.

Congestion control

NetScaler implements several enhancements to RFC3742 to improve TCP congestion management mechanisms. When NetScaler detects that there is a potential of the current transmission to cause congestion on the link, it places additional limits on the increase of the congestion window and the number of packets transmitted at a given time. This reduces the packet loss experienced by a flow because of congestion in the network and increases throughput.

Improved link utilization through intelligent packet retransmit

If NetScaler receives TCP packets that are out of order, it automatically predicts whether this is a result of packet loss or due to simple reordering. When it detects reordering, it assumes that the gaps will be filled quickly as missing packets arrive. This prevents the sending of duplicate acknowledgements, thus greatly reducing the number of packets transmitted and better optimizing the use of limited capacity WAN links.