ADC

Application switching and traffic management features

Below are the application switching and traffic management features.

SSL Offloading

Transparently offloads SSL encryption and decryption from web servers, freeing server resources to service content requests. SSL places a heavy burden on an application’s performance and can render many optimization measures ineffective. SSL offload and acceleration allow all the benefits of Citrix Request Switching technology to be applied to SSL traffic, ensuring secure delivery of web applications without degrading end-user performance.

For more information, see SSL offload and acceleration.

Access Control Lists

Compares incoming packets to Access Control Lists (ACLs). If a packet matches an ACL rule, the action specified in the rule is applied to the packet. Otherwise, the default action (ALLOW) is applied and the packet is processed normally. For the appliance to compare incoming packets to the ACLs, you have to apply the ACLs. All ACLs are enabled by default, but you have to apply them in order for the Citrix ADC appliance to compare incoming packets against them. If an ACL is not required to be a part of the lookup table, but still needs to be retained in the configuration, it should be disabled before the ACLs are applied. An ADC appliance does not compare incoming packets to disabled ACLs.

For more information, see Access Control List.

Load Balancing

Load balancing decisions are based on a variety of algorithms, including round robin, least connections, weighted least bandwidth, weighted least packets, minimum response time, and hashing based on URL, domain source IP, or destination IP. Both the TCP and UDP protocols are supported, so the Citrix ADC appliance can load balance all traffic that uses those protocols as the underlying carrier (for example, HTTP, HTTPS, UDP, DNS, NNTP, and general firewall traffic). In addition, the ADC appliance can maintain session persistence based on source IP, cookie, server, group, or SSL session. It allows users to apply custom Extended Content Verification (ECV) to servers, caches, firewalls and other infrastructure devices to ensure that these systems are functioning properly and are providing the right content to users. It can also perform health checks using ping, TCP, or HTTP URL, and the user can create monitors based on Perl scripts. To provide high-scale WAN optimization, the CloudBridge appliances deployed at data centers can be load balanced through Citrix ADC appliances. The bandwidth and number of concurrent sessions can be improved significantly.

For more information, see Load Balancing.

Traffic Domains

Traffic domains provide a way to create logical ADC partitions within a single Citrix ADC appliance. They enable you to segment network traffic for different applications. You can use traffic domains to create multiple isolated environments whose resources do not interact with each other. An application belonging to a specific traffic domain communicates only with entities, and processes traffic, within that domain. Traffic belonging to one traffic domain cannot cross the boundary of another traffic domain. Therefore, you can use duplicate IP addresses on the appliance as long as an addresses is not duplicated within the same domain.

For more information, see Traffic Domains.

Network Address Translation

Network address translation (NAT) involves modification of the source and/or destination IP addresses, and/or the TCP/UDP port numbers, of IP packets that pass through the Citrix ADC appliance. Enabling NAT on the appliance enhances the security of your private network, and protects it from a public network such as the Internet, by modifying your network’s source IP addresses when data passes through the Citrix ADC appliance.

The Citrix ADC appliance supports the following types of network address translation:

INAT: In Inbound NAT (INAT), an IP address (usually public) configured on the Citrix ADC appliance listens to connection requests on behalf of a server. For a request packet received by the appliance on a public IP address, the ADC replaces the destination IP address with the private IP address of the server. In other words, the appliance acts as a proxy between clients and the server. INAT configuration involves INAT rules, which define a 1:1 relationship between the IP address on the Citrix ADC appliance and the IP address of the server.

RNAT: In Reverse Network Address Translation (RNAT), for a session initiated by a server, the Citrix ADC appliance replaces the source IP address in the packets generated by the server with an IP address (type SNIP) configured on the appliance. The appliance thereby prevents exposure of the server’s IP address in any of the packets generated by the server. An RNAT configuration involves an RNAT rule, which specifies a condition. The appliance performs RNAT processing on those packets that match the condition.

Stateless NAT46 Translation: Stateless NAT46 enables communication between IPv4 and IPv6 networks, by way of IPv4 to IPv6 packet translation and vice versa, without maintaining any session information on the Citrix ADC appliance. A stateless NAT46 configuration involves an IPv4-IPv6 INAT rule and an NAT46 IPv6 prefix.

Stateful NAT64 Translation: The stateful NAT64 feature enables communication between IPv4 clients and IPv6 servers through IPv6 to IPv4 packet translation, and vice versa, while maintaining session information on the Citrix ADC appliance. A stateful NAT64 configuration involves an NAT64 rule and an NAT64 IPv6 prefix.

For more information, see Configuring Network Address Translation.

Multipath TCP Support

Citrix ADC appliances support Multipath TCP (MPTCP). MPTCP is a TCP/IP protocol extension that identifies and uses multiple paths available between hosts to maintain the TCP session. You must enable MPTCP on a TCP profile and bind it to a virtual server. When MPTCP is enabled, the virtual server functions as an MPTCP gateway and converts MPTCP connections with the clients to TCP connections that it maintains with the servers.

For more information, see MPTCP (Multi-Path TCP).

Content Switching

Determines the server to which to send the request on the basis of configured content switching policies. Policy rules can be based on the IP address, URL, and HTTP headers. This allows switching decisions to be based on user and device characteristics such as who the user is, what type of agent is being used, and what content the user requested.

For more information, see Content Switching.

Global Server Load Balancing (GSLB)

Extends the traffic management capabilities of a NetScaler to include distributed Internet sites and global enterprises. Whether installations are spread across multiple network locations or multiple clusters in a single location, the NetScaler maintains availability and distributes traffic across them. It makes intelligent DNS decisions to prevent users from being sent to a site that is down or overloaded. When the proximity-based GSLB method is enabled, the NetScaler can make load balancing decisions based on the proximity of the client’s local DNS server (LDNS) in relation to different sites. The main benefit of the proximity-based GSLB method is faster response time resulting from the selection of the closest available site.

For more information, see Global Server Load Balancing.

Dynamic Routing

Enables routers to obtain topology information, routes, and IP addresses from neighboring routers automatically. When dynamic routing is enabled, the corresponding routing process listens to route updates and advertises routes. The routing processes can also be placed in passive mode. Routing protocols enable an upstream router to load balance traffic to identical virtual servers hosted on two standalone NetScaler units using the Equal Cost Multipath technique.

For more information, see Configuring Dynamic Routes.

Load balances multiple WAN links and provides link failover, further optimizing network performance and ensuring business continuity. Ensures that network connections remain highly available, by applying intelligent traffic control and health checks to distribute traffic efficiently across upstream routers. Identifies the best WAN link to route both incoming and outbound traffic based on policies and network conditions, and protects applications against WAN or Internet link failure by providing rapid fault detection and failover.

For more information, see Link Load Balancing.

TCP Optimization

You can use TCP profiles to optimize TCP traffic. TCP profiles define the way that NetScaler virtual servers process TCP traffic. Administrators can use the built-in TCP profiles or configure custom profiles. After defining a TCP profile, you can bind it to a single virtual server or to multiple virtual servers.

Some of the key optimization features that can be enabled by TCP profiles are:

  • TCP keep-alive—Checks the operational status of the peers at specified time intervals to prevent the link from being broken.
  • Selective Acknowledgment (SACK)— Improves the performance of data transmission, especially in long fat networks (LFNs).
  • TCP window scaling— Allows efficient transfer of data over long fat networks (LFNs).

For more information on TCP Profiles, see Configuring TCP Profiles.

CloudBridge Connector

The Citrix NetScaler CloudBridge Connector feature, a fundamental part of the Citrix OpenCloud framework, is a tool used to build a cloud-extended data center. The OpenCloud Bridge enables you to connect one or more Citrix ADC appliances or NetScaler virtual appliances on the cloud-to your network without reconfiguring your network. Cloud hosted applications appear as though they are running on one contiguous enterprise network. The primary purpose of the OpenCloud Bridge is to enable companies to move their applications to the cloud while reducing costs and the risk of application failure. In addition, the OpenCloud Bridge increases network security in cloud environments. An OpenCloud Bridge is a Layer-2 network bridge that connects a Citrix ADC appliance or NetScaler virtual appliance on a cloud instance to a Citrix ADC appliance or NetScaler virtual appliance on your LAN. The connection is made through a tunnel that uses the Generic Routing Encapsulation (GRE) protocol. The GRE protocol provides a mechanism for encapsulating packets from a wide variety of network protocols to be forwarded over another protocol. Then Internet Protocol security (IPsec) protocol suite is used to secure the communication between the peers in the OpenCloud Bridge.

For more information, see CloudBridge.

DataStream

The NetScaler DataStream feature provides an intelligent mechanism for request switching at the database layer by distributing requests on the basis of the SQL query being sent.

When deployed in front of database servers, a NetScaler ensures optimal distribution of traffic from the application servers and Web servers. Administrators can segment traffic according to information in the SQL query and on the basis of database names, user names, character sets, and packet size.

You can configure load balancing to switch requests according to load balancing algorithms, or you can elaborate the switching criteria by configuring content switching to make a decision based on SQL query parameters, such as user name, database names, and command parameters. You can further configure monitors to track the states of database servers.

The advanced policy infrastructure on the Citrix ADC appliance includes expressions that you can use to evaluate and process the requests. The advanced expressions evaluate traffic associated with MySQL database servers. You can use request-based expressions (expressions that begin with MYSQL.CLIENT and MYSQL.REQ) in advanced policies to make request switching decisions at the content switching virtual server bind point and response-based expressions (expressions that begin with MYSQL.RES) to evaluate server responses to user-configured health monitors.

Note: DataStream is supported for MySQL and MS SQL databases.

For more information, see DataStream.

Application switching and traffic management features