SANs: Enabling EAI and eBusiness
By Fred Moore 101 Storage Solutions
Considering the market visibility given to the topic, most of us have, by now, read about or perhaps even been involved in implementing a SAN (storage area network). Can we, however, say the same about EAI (enterprise application integration)? This article takes a look at these emerging technologies to better understand their increasing, reciprocal value.
EAI Middleware Overview
Relatively new in the marketplace, EAI is packaged software that provides the foundation for integrating enterprise software applications, databases, and business processes that are running on different computer systems--it's the middleware glue that helps it all work together. EAI solutions significantly reduce the time and cost needed to develop, integrate, and maintain applications existing on legacy, open systems, and distributed platforms. Application integration thus includes any effort that enables two or more application systems to communicate and share each other's most valuable assetæthe data, and as such is a key enabler for many eBusiness applications.
Analyst projections call for the EAI middleware revenue market to exceed $5 billion by year-end 2003, a CAGR of more than 75 percent over 1999's $600 million market. The EAI market is being driven primarily by the
(1) re-centralization of computing resources,
(2) need to access data from applications running on OS/390 (z/OS), Unix, Windows NT, and Linux platforms,
(3) tremendous effort and expense required to integrate applications through labor-intensive point-to-point programming and data migration or conversion methods, and
(4) rapid proliferation of eBusiness and Internet applications.
Today many customers, who are considering EAI, already have significant investments in batch-oriented systems. However, modifying such systems is difficult given the well-publicized shortage of software developers. Costs of modifying existing software applications are estimated to range from $2.50 to $4.00 per line of code!
In the late 1980s and early 1990s, several trends led to the development of distributed client/server computing systems, which attempted to break down what was perceived as being larger, more costly centralized computing environments. Applications developed in various business areas were seldom coordinated. As a result, islands of information were created with little thought given to the longer-term requirement to communicate and share information between these islands. Such islands were often built around different computing systems, with Unix, Windows NT, and OS/390 (the traditional mainframe) at the core.
However, by the mid-1990s, we had come to realize that distributed computing and its associated management requirements often created redundancy and was significantly more expensive and labor-intensive than managing resources centrally. This reality has fueled the present trend toward centralizing computing resources, while distributing access. In addition, changing applications in these systems had been difficult and costly, a situation that isn't likely to change in the near future, given the high availability characteristics of such systems (now at more than 99 percent for Unix and 99.999 percent for the OS/390 Sysplex).
Centralized Management and Distributed Access
Enterprise class systems continue to grow in number as the requirements for ultra-high availability and scalability increase on a near vertical trajectory. Nonetheless, much of the tremendous IT growth is based on open systems platforms, typically Windows NT, Linux, or Unix platforms, as it's easier to develop new applications on these systems.
Unix and NT together account for more than 60 percent of the annual disk storage market revenues. Open systems, with their inherent support for and standardization of TCP/IP networking, low-entry cost, and ease-of-implementation, are excellent choices for implementing the distributed architectures that are used throughout eBusiness applications.
What happens when data movement and shared data access becomes more intensive between intra-business departments or between businesses (B2B) and outgrows the I/O capabilities of LAN-based data transfer? Unfortunately, the goal of real-time universal data access has, until now, been basically unattainable.
Even though different implementations achieve part of the goal, each misses in at least one critical criterion: Either the access isn't instantaneous or the data isn't in its native format. Or, the system can't directly access data where it resides on the network, and as a result must move the data from the core database, frequently, reformatting the entire file. Such solutions are expensive, proprietary, time-consuming, and resource-intensive to implement. The need for a true data sharing storage architecture going forward is obvious.
Data Sharing and SANs
Does the growing EAI and eBusiness market clearly identify an application area that SAN architecture has targeted? If one server didn't need to access information on another server, EAI and even eBusiness markets wouldn't exist.
Our highly integrated world of computing suggests that connecting all types of nodes on a worldwide network is no longer an option. The EAI model of connecting unlike servers and their respective users to enable data access and data sharing nicely overlays the long-range goals of the SAN model for allowing similar and heterogeneous servers to access a common pool of storage.
The SAN may be ahead of its time, but is not ahead of its vision. Keep in mind that NAS provides a good solution in low- to moderate-level data intensive systems. The longer view for SANs significantly challenges and will eventually replace the traditional practice of using a general-purpose server and dedicated storage as a storage repository. The high-speed network fabric constructed from high-speed interfaces connects a pool of data storage devices to a group of servers, allowing storage devices and servers to be added and removed independently of each other. This vastly improves availability over existing monolithic EAI storage structures.
In the SAN, data is shared rather than connected to and owned by a server. An unscheduled outage on a server need not affect access to data in the SAN model. Emerging SAN management software, virtualization techniques, and data sharing software solutions are being developed that enable and enhance the trend toward EAI. As EAI applications demand larger data intensive solutions, SANs enable application servers to integrate data in legacy systems, databases, and existing core applications with newly created Web applications, ERP, CRM, and the now exploding category of eCommerce and eBusiness applications.
As Web and eCommerce applications continue to grow in size, complexity, transaction volumes, and importance, the need for ease of management in designing, developing, and deploying them grows as well. A report from Solomon Smith Barney "The SAN Book" states that Amazon.com consumed 42TB in six months and Mail.com filled 28TB in 45 days! Twenty-eight terabytes approximates the total traffic per month on the Internet just three years ago. Such applications don't respond to any known capacity planning methodology!
The coupling of these requirements with the premise that SANs significantly reduce the complexities and costs associated with managing storage further highlights the importance of SANs to future EAI needs. Most of the large operating system vendors are beginning to provide an end-to-end method for integrating legacy systems with distributed platform applications and are actively partnering with key, strategic middleware and data sharing providers.
Final Thoughts...
Even before the SAN concept has been implemented on a widespread basis, we can see that the new wave of application requirements and their potential storage service providers, present even more extensible uses for the SAN. With Web users now predicted to exceed 500 million people in three to four years, and eCommerce and eBusiness revenues expected to surpass a trillion dollars in the same timeframe, growing a more functional storage infrastructure is more critical than ever.
101com.com |
