|
The LoadMaster resource monitoring provides data on both server memory and CPU, ensuring users experience the most efficient load balancing possible across each server. LoadMaster integrates seamlessly with WTS Session Directory – providing a reliable “re-connect” when a remote desktop connection to the server has disconnected. LoadMaster also provides RDP-based Layer 7 persistence that incorporates client session reconnect, which can be utilized without the need for the Session Directory service to be installed. This helps simplify IT infrastructure, and provides cost savings benefits.
LoadMaster enables WTS users to maintain persistence, as well as perform resource monitoring for Window servers running multiple services. The LoadMaster resource monitoring feature provides data on both server memory and CPU, ensuring users experience the most efficient load balancing possible across each server. Additional benefits from LoadMaster load balancing WTS include, service checking for servers running Microsoft WTS, and extended capabilities for the increasing segment of remote workers.
Application Delivery Controllers
Application delivery solutions were built to address the challenges associated with website infrastructure complexity, performance, scalability and security. They may be known as application delivery controllers (ADC), application delivery controllers, server load balancers (SLB), application front-end devices (AFE), application traffic managers, and web front-ends, content switches and application switches. In order to avoid confusion, this paper will focus on datacenter solutions, and refer to application delivery solutions as application delivery controllers. Today’s application delivery controllers actually evolved from server load balancers that were first introduced in the late 1990s.
ADCs provide the ability to direct Internet users to the best performing, most accessible servers. Should one of the servers (or applications on that server) become inaccessible due to any type of failure, the ADC will take that server off-line, while automatically re-routing users to other functioning servers. This process is essentially seamless to the user, and critical to servicing the customer.
During the past five years, application delivery has emerged as one of the most important technologies in solving the problem of performance and accessibility for Internet-based applications.
In addition, by using various load balancing algorithms, an ADC can distribute users to servers that offer the best possible performance. The ADC can dynamically interrogate key server elements such as the number of concurrent connections and CPU/memory utilization.
To further enhance, and secure the user experience, more-advanced ADCs provide SSL offload/acceleration. SSL acceleration in the ADC enables you to offload the SSL handshake and encryption/decryption processes from the servers. This offloading dramatically increases the servers’ performance, while decreasing the time and costs associated with the server’s SSL certificate management.
The application delivery market is divided among two ADC vendor groups. High-end vendors such as F5, Cisco, and others which often cost over $100,000, after you get the redundancy and options you need; and the "value" ADCs vendors such as Kemp, where a full high-availability pair of loadmaster products start at under $5000.
KEMP LoadMaster’s unique WTS support
Kemp Technologies LoadMaster supports any TPC or UDP port/protocol combinations. Before LoadMaster, if you wanted to do connection load balancing based on more than a simple number of current network connections, you had to write a script that calculated the load of each host that was accessible via HTTP by the load balancers. The script ran on each host, and generated a number from 1 to 100. Then the load balancer would check http://host1/myscript, http://host2/myscript to determine which server should receive the new load.
This was not the most ideal solution, but it was flexible, since you could use any perfmon counter, WMI interface, EXE, etc. to generate the numbers. For example, you could create a script that ran several checks for websites, and sample database queries. If everything came back as expected, it would generate a load number. If any aspect failed, the script generated a value which would tell the load balancer that server 1 was "full," thus preventing new connections to the broken server.
You could configure them to balance user traffic to Terminal Servers and listen to port 3389, and even install IIS on your servers to create load balancing algorithms. This worked except for one thing; if a user with an existing disconnected session reconnects, there is no way for the load balancer to know about that existing disconnected session, so the user is routed to the least-busy server. Chances are they'll be sent to a different server, thus starting a new session even though they already had an existing session.
|
