Datacentres are going through various stages of
consolidation. Driven by the development of technology and the
need to contain costs and improve services, the system
infrastructure has become more resilient, more available and more
secure.
To achieve these goals, servers and storage have to be
centralised or pooled. This is the first phase of consolidation.
Once it has been completed, the business benefits from a
rationalisation of the system infrastructure to allow IT
departments to better manage the overall resources.
Storage networking has played a significant role in this
change. An aspect of this first-phase consolidation is that
systems tend to be supplier-specific.
Research by Macarthur Stroud International revealed that by the
end of 2004 most organisations had consolidated their storage and
servers, 66% had consolidated both, with the other 34%
consolidating either storage or servers.
But there are many applications and environments that must be
supported over daily, weekly and monthly system cycles.
To meet this requirement, infrastructure services are being
developed to better use system resources. Technologies such as
VMWare and Infiniband are supporting server virtualisation, and
IBM's San Volume Controller and EMC's Invista enable storage
virtualisation in the network.
Virtualisation tools are triggering the second phase of
consolidation. System management gains are enabling organisations
to increase the security and availability of their systems with a
sound return on investment. The business and system needs that are
being addressed are:
- Protecting and securing the data at different sites across the
organisation.
- Introducing security policies to permit access to approved
users and applications
- Resource management to address fluctuating workloads and to
enable system reconfiguration if system performance is impacted or
any failure occurs.
The visible aspect of this second phase of consolidation is the
sure and steady acceptance of virtualisation technologies.
Physically, this is visible from the consolidation of switches and
directors (managed switches) into larger directors within the
datacentre, while there are ongoing consolidation projects within
distributed business units and branch offices.
In the third phase, it is necessary to provide supporting
services across the network. Data protection practices are backing
up users with wide area file support applications and in
high-volume applications with the use of storage routers.
The need to have faster system restores means new processes need
to be put in place. Moving to a continuous data protection
environment built on a system infrastructure will enable fast and
rapid data movement.
The network management functions already define access routes
and permit approved access. Tracking the activity at each node in a
network will become of increasing importance as system security is
enhanced.
Managing the IT infrastructure in more automated ways will also
be an area that system managers will address. The service level
expectations of users are that systems are always available.
Although delivering the service level agreements may vary from
business-critical processes to support applications, the impact on
each and every user is additional cost, if systems on which they
depend are not available.
Resource management tools not only track the utilisation of
server, storage and network resources, they enable the
reconfiguration of systems by preset policies or through management
direction.
Finally, the implementation of high-speed storage networks,
whether Fibre Channel or IP, is setting the framework by which the
next phase of system productivity developments will be realised.
Virtualisation, across all system and network components, is
driving the change.
Hamish Macarthur is founder of IT analyst firm Macarthur
Stroud International