Application Delivery (ADX)

Application Availability and Scalability

by pmorrissey on ‎04-30-2009 03:47 PM - edited on ‎10-30-2013 05:46 PM by bcm1 (888 Views)

Why is it important...Draft


Today, IP applications are at the heart of business operations, and these  applications are used for various business-critical services that help reduce  cost, increase productivity, and grow profitability. Downtime associated with  these critical applications has a direct negative financial impact on the  organizations. The cost of downtime keeps growing as reliance on  mission-critical applications grows. Many mission-critical application  transactions are time sensitive, and even poor response time, if not downright  failures, could have severe negative impact on the end users of these  applications and the business operations. For example, imagine a financial trade  getting executed few seconds late because of poor application performance, or a  request for patient information delayed during a medical emergency. Maximum  uptime and peak performance are the critical requirements that must be met.


Increased demand for application resources results from growing number of  users, increased volume of application transactions, changes in the nature of  application usage, and implementation of new services. And, there is no  indication that the need to scale applications to meet growing demands will ease  soon. Accommodating growth and never-ending changes to applications requires  flexible application and server farm infrastructure that can scale on demand. A  critical component to growth is growing transparently to existing application  users. Virtual infrastructure decouples physical resources from application  availability and user capacity. It eliminates the need for forklift upgrades  that destroy invested capital as business requirements change. Protecting  invested capital is the first step to maximizing Return on Investment (ROI), and  growing on demand conserves capital, which further maximizes the ROI. The  virtual infrastructure also keeps application availability at a maximum  independent of the physical changes to the infrastructure.


Brocade ServerIron high-performance application switches provide maximum  availability to applications by intelligently distributing traffic among all  available servers, and dynamically monitoring the ability of servers and  applications running on the servers to deliver optimal performance. Using  customizable health checks at various levels of granularity like host, port,  application and transaction levels, load balancers transparently react in real  time to increases and decreases in available server capacity by redistributing  client traffic. Failures of servers and applications running on the servers do  not cause disruption to the end clients. The failed resources are simply set  aside until they come back online, and client connections are re-directed to  other available servers. ServerIron switches can be deployed in multiple  high-availability modes with stateful session synchronization and failover to  preserve client connections during catastrophic failures. Applications are  transparent to the presence and failures of load balancers, and function  uninterrupted.

Scaling applications and server farms is a fundamental requirement for  continued business growth, and is cost-effectively and transparently met by the  ServerIron application switching solution. ServerIron switches provide virtually  unlimited scalability to IP-based applications by allowing the use of multiple  servers with load balancing and failover. The switches allow businesses to  leverage commodity servers to build highly sophisticated and secure application  infrastructure. As demand grows, simply slide in new server resources and  configure the ServerIron switch to use the new resources. There is no need for  forklift upgrades to the server farms and disruption to applications. With the  fine-grain content switching intelligence of these switches, there is no longer  the need to scale the entire server farm simultaneously. Growth can be targeted  to specific application functions and content servers, which allows even higher  levels of scalability. Additionally, existing resources can be more efficiently  used by implementing content switching rules that maximize overall  utilization.