The Challenges and Perspectives

Purely from an information systems perspective, the datacenter is a
heterogeneous mix of servers, operating systems, storage and management tools.
Along with an increase in technology complexity, the emerging trends in these
areas pose significant challenges.

Hardware and operating systems trends, multi-tiered architectures, TCO, and
evolving platform solutions (blades, bricks, racks and grids) have a bearing on
how you decide to consolidate your servers and what goes into your datacenter.
With ever-growing data volumes, effective storage tiers that address storage
management, networked solutions, disk technology and enterprise backup, archival
plus compliance become the backbone of your IS infrastructure. Few industry
trends have generated more attention and debate than utility computing. The
definition and ultimate form of utility computing varies by the vendor, all
promising to fundamentally change the way enterprises deploy and consume
computing power.

The prospect of a datacenter

The increasing demand for storage capacity, coupled with the challenges of
increasing data accessibility and adhering to new compliance regulations for
data retention, has spurred organizations to reassess the structure of their IT
infrastructure.

There are several "possible" technologies on the horizon that have
an impact, not only on the efficiency and effectiveness of the datacenter, but
also the availability and resiliency of it. Commodity hardware, Linux and open
standards have the potential to dramatically decrease the technology risks that
companies are exposed to.

A solid information security infrastructure within the datacenter network is
just as important as it is within the local area network in the corporate
office, however, it's often overlooked or considered too expensive to address.

Even if you are considering outsour—cing your IT operations, there is a lot
that you need to assess about the capabilities of your service provider. It is
highly imperative to evolve the best method to determine, assess and review an
outsour—cing company's ability to recover or maintain availability.

Each square inch of datacenter space is prime real estate with a monetary
figure for not only the space, but for ongoing operations and support. And,
based upon your business needs, the tier level and redundancy of your datacenter
should provide protection against downtime; however, the budget must be balanced
to assure that unnecessary monetary expenditures are avoided. New datacenter
design considerations based on TIA/EIA TR942 Telecommunications Infrastruc`ture
Standard for Datacenters are the way out.

The biggest challenge facing datacenter designers and users today is
providing power and cooling to high-density equipment installations. The rapid
emergence of "1U" and "Blade" servers, and the ever
increasing demand for high-volume storage, present power and cooling demands
that were beyond comprehension just a few years ago. Most frightening of all to
IT personnel is the specter of liquid-cooled cabinets in the datacenter.

Iishwar Daas

The Four Essential Elements of System Availability

The CIO should look for an infrastructure which is designed with a focus
on reducing the Mean Time Between Failures (MTBF) and, more importantly,
minimizing Mean Time To Recovery (MTTR). He should look at these elements for a
comprehensive availability strategy

Power Continuance Since power problems are the largest single cause of
downtime and data loss (Contingency Planning Research), a fail-proof UPS
strategy is one of the first steps a company needs to take to ensure system
availability. Look for vendors who can offer:

wide range of UPS from single 300VA to parallel redundant 1MW systems,
shutdown software for applications and servers to minimise loss and restart time
and networked managed UPS capability.

Power Distribution Messy, tangled power cables with multiple cascading
power distribution units can significantly delay the time to isolate a problem
and return power. Advanced power distribution minimises downtime due to operator
error. Look for vendors who can offer rack-mount power distribution units with
built in current metering and alarms, remote individually controlled power
outlets for power cycling, reboot and shutdown sequencing and automatic transfer
switches provide redundant power to connected equipment.

Secure Physical Environment In today's networks, the physical
environment is critical to the availability, security and reliability of your
equipment. Today's IT systems are extremely susceptible to environmental
problems. Racks filled with servers and other equipment can produce a lot of
heat, which creates the potential for data corruption. A secure physical
environment is made up of-rack systems to house all your IT equipment, thermal
management to cope with high power densities (over 5kW for a 42U rack), wide
range of racks for datacenter and wiring closet applications (13U, 25U, 40U, 42U
and 47U), etc

Integrated Manageability Intelligent network managed devices and
software drastically reduce Mean Time To Recovery (MTTR-includes the time it
takes to reboot the computer, restart applications and reconnect to the network)
by providing visualization and exception management.