Accelerate The Enterprise

Accelerate The Enterprise

June 26th, 2006:Server centralisation may appear to be one solution to remote office integration. But it represents issues of control and cost itself. Shaun Page, Vice President of Juniper Networks, ANZ gives his opinion on where WAN acceleration should go.

Server centralisation has historically failed to address the issue of performance. By collating the entire organisation's applications and data storage services into discrete and centralised sites, a company can reduce the number of staff and the time and effort needed to manage and protect the collection of servers. It can also minimise space, power and connectivity costs, as compared to using servers laid ad hoc across multiple branch offices. Yet this centralised arrangement of a "star" topology – where users all connect directly to one, or a mere handful, of central servers – also means centralised demand on the finite bandwidth provisioned at these "data centres."

Justice
Shaun Page, Vice President of Juniper Networks: "Integration of WAN and data centre optimisation solutions allows acceleration of the widest cross-section of business applications, improve delivery options..."

Budget Cuts
Facing limited budgets, IT administrators have little choice but to implement the cost savings measures of centralisation, especially when also burdened with the task to comply with business regulations – corporate governance laws created in the aftermath of spectacular financial collapses including Enron and natural disasters such as the Tsunami of 2004.

These regulatory requirements include data redundancy and other continuity measures to protect mission critical information and business infrastructure, and data retention policies such as long term email archival for record retrieval.

Comprehensive network security is also a prime impetus for server centralisation. In today's environment where multiple forms of attack are originating from multiple sources – including organisational insiders whether deliberate or not – it is easier to protect and repair infrastructure when servers are deployed in as few locations as possible. Where fault occurs, either because of network intrusion or as a result of other hardware or software failure, troubleshooting is simplified when IT response teams have fewer machines to examine.

Although centralised data storage and access is the only logical solution given all the aforementioned criteria, there has not – until recently – been an elegant solution to the performance issue. The historical "brute force" method of a stop-gap fix to the performance problem is expensive and simplistic: just buy more bandwidth from the provider. It's expensive because the additional bandwidth needed to provide a satisfactorily responsive experience for all the connected users often far exceeds assumptions. It's simplistic because it does not take into consideration the constantly changing nature of network traffic – especially when the infrastructure needs to support latency-sensitive and bandwidth-hungry services such as streaming video, or real-time financial transactions.

Fortunately, there's a better way. Application acceleration solutions are now available and mitigate many of the problems created by server centralisation, without necessitating the historically expensive answer of buying more bandwidth. Acceleration solutions will also maximize return on investment in bandwidth, provide easier network administration and troubleshooting, and increase traffic performance even with realtime packet services. Application acceleration solutions perform their magic by unifying "Web tier" acceleration and WAN acceleration solutions. These include all the load balancing, data compression, caching, TCP multiplexing, SSL protocol acceleration, network and protocol security methods used as to squeeze more productivity from servers and WAN connections. What's different now is the new acceleration architecture consolidates all the disparate technologies into a seamless and cooperative model. The new acceleration architecture uses works best, minimising the load on server clusters by intercepting remote requests and making dynamic decisions on managing these requests. Acceleration effectively increases server capacity, often doubling available capacity on existing servers and cutting page download times in half.

The organisations that ultimately succeed in finding an effective knowledge management solution arent those that necessarily use the best software or design the best processes, although both aspects are critical to success. Its the organisations that simplify the knowledge management process by making it easier to collect, store, share information and make it part of the daily business processes that come up tops every time.

So, how do you go about encouraging what is, essentially, a cultural change in your organisation?

Billion Dollar Brains

As enterprises embark on server centralisation plans – against the backdrop of business regulation, overhead minimisation and stronger security – application acceleration will emerge as a crucial enabler bridging the benefits of centralisation and user experience. Integration of WAN and data centre optimisation solutions allows acceleration of the widest cross-section of business applications, improve delivery options, and providing unparalleled monitoring and reporting tools to track performance. And the market is quickly realising their benefits.

Consulting firm Frost & Sullivan has estimated a rapid 30 per cent compound annual growth rate (CAGR) for the Asia Pacific application front end acceleration market (2004 to 2011), becoming a billion dollar industry by the end of that estimated period.

Business Solution: