Infrastructures
and Applications
In a typical neighborhood, people live on a thin veneer of essential services that are for the most part hidden from view. Power lines, phone lines, water pipes, gas and sewer lines are thankfully buried
underground
. This collective plumbing provides the support infrastructure for daily living and becomes a focal point only when things don't work and daily living is disrupted.
This same scenario applies to the typical neighborhood of data communications. End users run a variety of applications to
transact
their daily business,
assuming
that the underlying data communications network will provide stable and continuous service. If the network goes down, daily business is disrupted, sometimes costing companies millions of dollars per
hour
. End users are not and should not be aware of the underlying plumbing that is so essential to their business lives. That is the responsibility of network managers and
engineers
who work behind the scenes to ensure that the upper layer applications run smoothly.
Database applications, online transaction processing, email, web services and other end
user
applications require computing platforms, storage and network interconnect. Storage area networking
brings
the storage component into the network plumbing and enables sharing of disk and tape resources by all upper layer applications. Since storage networks are a relatively new technology, there has been more initial focus on SAN infrastructures. Fibre Channel, iSCSI, IP-based SAN extension, storage virtualization and other plumbing issues receive considerable press coverage and engender heated debate among SAN
vendors
. While this may be a natural side effect of an emerging technology, in the end all the customer really wants is reliable and high performance data service: Will my applications run more
efficiently
and will I have continuous access to my data?
Specific storage-
related
applications such as volume managers, storage resource management, server clustering, data replication, and tape backup utilities benefit from the lower level storage network's performance and resource sharing capabilities. Those applications, however, should be
agnostic
to the underlying transport. The customer should expect that, regardless of whether the SAN infrastructure is based on Fibre Channel, iSCSI or some combination of both, that the upper layer applications will perform the same. One of the challenges of SAN hardware vendors, then, is to make their solutions transparent to the higher level storage applications. The ideal network solution is one that, like an underground water main, is invisible to the end user.
The stability and interoperability of SAN infrastructures is fostered by standardization. To maintain a low profile and minimize administrative overhead, SAN hardware needs to
satisfy
the customer's expectations for standards compliance. Consequently, customer hardware acquisitions through Requests for Proposals (RFPs) typically have long laundry lists of standards checkboxes that must be filled before a solution is
considered
. The assurance of standards compliance
increases
the customer's odds that an acquisition will not later create support and interoperability issues. The idea here is to verify that, once in place, the plumbing will pose minimal issues.
Ironically, these same criteria are not placed on the applications themselves. While running applications smoothly is predicated on a stable, standards-compliant network infrastructure, the applications
themselves
are vendor-specific or tailored for a customer's business needs. Applications thus benefit from the rigor taken to create sound plumbing, but can provide the flexibility to adapt to specific customer requirements.
|
