CONVERGED INFRASTRUCTURE or ‘unified computing’ is essentially an ‘in-a-box’ approach which relies on a specific vendor supplying pre-configured packages of hardware and software in a single stack. This includes servers, compute, storage and network resources which are shared by numerous applications. These are bundled within a single framework that can be managed collectively and centrally by policy-driven procedures. The main goal for organisations with consolidated systems is to maximise utilisation.
If carried out successfully, the transition to converged infrastructure can offer many benefits. However, there are also a number of potential obstacles and uncertainties to be aware of that if not addressed could cause considerable inefficiencies and undermine a company’s consolidation strategy:
Vendor Lock-In
In order to realise the benefits of CI, most converged offerings will have strict configuration guidelines that must be adhered to. So once entered into, the end user must buy from the same manufacturer to remain in compliance with the specification. Should they need to add a server, switch ports, or storage for their systems to scale, they should be aware that there is often no room for procurement outside of the CI vendor offering.
Migration performance uncertainty
With all of the variables and fluctuating workloads within a typical data centre, if any non-converged application infrastructure is already performing inadequately, how can IT decisions makers be assured that moving an application to a converged infrastructure system will not negatively affect its integrity and performance? How can they really be sure that this migration will bring improvement and not make things even worse?
Application resource sharing
Since there is usually only one shared pool of resources for each ‘pod,’ how can application owners be sure that another application won’t starve their application of resources? Without the visibility of shared workload resourcing, how can IT management avoid issues or even identify where they exist? QoS policies can only cover the application itself and cannot be held accountable for how it’s affected by other applications competing for the same resources.
Limited Management Integration
To ensure the best possible CI performance organisations need to gain an accurate understanding of the many connections within the converged environment. But obtaining useful performance metrics can be difficult to achieve because although unified management interfaces are provided as part of CI solutions to manage the consolidated stack, they are proprietary, lack correlation, advanced analytics and the reporting capabilities for contextual integration across the server, network and storage resources. In addition, performance inconsistencies can continue to exist due to legacy tools still used to manage their existing non-converged open systems environment.
What this demonstrates is that even within the promise of the ‘simplified’ converged infrastructure, fragmentation and IT sprawl on a granular level still exists. So for the end user, if not managed well, convergence can also result in a highly complex infrastructure with increased vulnerability to problems and component failures. It is the unpredictability and lack of visibility prevailing at this level which is driving the need for advanced analytics, which must be in real-time to accurately report on performance.
How Infrastructure Performance Management strengthens the converged environment
As with CI adoption, IT decision makers are always looking for ways to optimise and consolidate their data centres, reduce costs and improve efficiencies. The answer, or part of it, comes in the form of Infrastructure Performance Management (IPM), a new and vital component in datacentres that allows IT administrators to gain an end-to-end view of the entire IT infrastructure by enabling them to continuously capture, correlate and analyse system-wide performance, utilisation and health, in real-time.
The results of a survey of European IT decision makers by Virtual Instruments in November 2014 revealed that 86 per cent of respondents agreed that IPM could help improve the IT department’s performance in terms of meeting the needs of the business.
The Virtual Instruments solution, VirtualWisdom4® is the industry’s first Infrastructure Performance Management (IPM) platform, built from the ground up to unify the monitoring of dynamic converged infrastructures and traditional data centre resources.
VirtualWisdom4® is the only solution with the resources and analytics capability to correlate information from every tier of the stack to provide unique insights about the performance and any issues going on throughout the entire infrastructure, end to end, regardless of vendor.
VirtualWisdom4® ensures performance-
based SLA commitments are still
maintained once applications are moved
into a converged infrastructure, allowing
more applications to be virtualised
while checking that the capacity of the
physical server is used to its maximum
potential resulting in reduced OPEX and
CAPEX.
The solution enables end-to-end visibility
to track and monitor performance from
the application through the converged
infrastructure to pinpoint and resolve problems before they become serious,
protecting critical application performance
and investments.
Since VirtualWisdom4® is a vendor-
agnostic protocol based IPM platform
(now with a 16G Fibre Channel line-rate
capability), it not only works with several of
the major converged systems offerings,
but it can also be used to monitor existing
legacy open systems infrastructure. So
it provides a management solution that
can baseline current system performance
while also tracking and monitoring the
new converged deployments to increase
operational efficiency - allowing the
flexibility to adopt new technologies as
required.
The VirtualWisdom4® IPM platform fully supports converged environments by providing detailed visibility into the performance, utilisation, and health of the infrastructure. With this, customers can now not only improve the performance, but also the availability and cost efficiency of their end-to-end infrastructure.
So while it is clear that converged infrastructures can improve overall agility
and efficiency and remove many of the issues associated with open systems integrations, there is no doubt that some of its limitations create new challenges
that must be addressed in order to
ensure that the convergence strategy is successful.
With flash, cloud migration, software
defined and convergence all combining to cause massive disruption in today’s datacentres, Infrastructure Performance Management (IPM) is the only way to guarantee optimum performance in
mission-critical applications supporting
the business.
