The application economy places unprecedented demands on businesses and ICT to deliver engaging services, improve time to market and provide access to information, anytime, anywhere, on any device.
To transform and win in the application economy, a new network is needed. For example, AT&T is to virtualise 75% of its network by 2020. In terms of service providers and large enterprises - if they can't transform, their customers are likely to go elsewhere.
This is the beginning of a revolution in today's networks, which is being driven by software-defined networking (SDN) and network functions virtualisation (NFV). However, technology gaps still exist that will derail adoption and deployment initiatives.
The application economy exerts pressure on both the data centre and the network in support of service agility and service velocity. With the proliferation of virtualisation and cloud computing comes the agility in the data centre, which is crucial to provision the service velocity required for success in a world of digital transformation.
The next part of the agility required in the network space comes from what is delivered by SDN/NFV. Service velocity is the rate at which agility can be utilised to create and change new services while guaranteeing the quality of service demanded by today's customers.
Dynamic infrastructure
The current state of SDN/NFV is not unlike what was experienced some years ago, during the evolution of virtualisation in the server space in support of cloud computing. This development led to the enablement of automated resource allocation in line with business demand, thereby allowing organisations to maximise resource utilisation and sweat their investments in the data centre.
This, in turn, directly contributed to improved return on investment.
Until recently, the networks interconnecting and supporting these dynamic environments have remained relatively static. In the last several years, network virtualisation efforts have gained momentum, extending the agility in the server space to the rest of the resources supporting the service transaction path. While the network services have always been offered with stability, reliability and predictability in mind, these could be affected if order is not maintained along this road to service velocity.
Traditional monitoring tools can be rendered ineffective in the new agile network world. Now, with SDN/NFV, the configuration and set-up of the network resources supporting business can change quickly - so quickly that management solutions might encounter the problem of 'stale state'. This occurs when the management solution is monitoring an environment against an architecture with a set of key performance indicators that is not up to date with the changes that already occurred in that environment.
An example of this is the performance and availability monitoring thresholds set to monitor specific network resources supporting key customer services. When the service is changed, or a new service is added, especially at the rate needed for service velocity, the information contained in the management solution versus the resources supporting the service becomes out of sync.
Historically, most service providers have scheduled network discoveries to once a week, and lately, to once a day, as network changes become more frequent. But, even this is not acceptable for SDN/NFV environments, where hourly checks, even if they are possible, could be ineffective. Most network discoveries are executed utilising simple network management protocol (SNMP), which uses a polling paradigm, and as such, could result in excessive traffic overhead on the monitored resources, and in turn customer services.
The inability of legacy management solutions to cope with dynamic next-generation resources is exacerbated by the increasing scale of the network in order to provide for the application economy or digital transformation. This new fast-paced world, with users demanding access to information, anytime, anywhere and from any device, not to mention the ever expanding Internet of things (IOT), requires Web-scale computing.
IOT will increase the number of manageable objects by an order of magnitude.
IOT will increase the number of manageable objects by an order of magnitude. This will all lead to the requirement for agility at scale.
The need for service transaction insight across the entire transaction path becomes even more relevant, with a requirement for integrated application aware network performance monitoring. It also requires application performance management across an agile network service environment, where closed loop service life cycle management is required in conjunction with SDN and NFV orchestration and fulfilment.
Therefore, management solutions for service velocity in an SDN/NFV world need to cater for:
* Multi-vendor environments;
* SDN/NFV alongside traditionally fulfilled, assured resources and services;
* Complete awareness of the overlay and underlay relationships and dependencies, advanced scale and flexibility; and
* SNMP and non-SNMP physical and virtual resources (routers, switches, firewalls, network function servers).
Solution for success
To succeed, operations centres need to seize the opportunity to completely redefine their infrastructure management approach to be as agile as SDN/NFV demands. A solution is required that can bridge the existing management solution capabilities with the needs of SDN/NFV service velocity, reducing the complexity in network visibility and planning, while enabling self-service automated networks.
This solution must facilitate reduced complexity, troubleshooting and expert insight into the service transaction path, plus configuration across the SDN/NFV stack. The solution also needs to provide extended visibility for complete awareness into the overlay and underlay relationships and dependencies, advanced scale and flexibility required by highly dynamic SDN/NFV ecosystems.
With a management solution designed for the dynamic next-generation resources present in an SDN/NFV environment, telcos, service providers and enterprises alike will be able to leverage the agility inherent in these technologies, and gain the service velocity needed to cope with and thrive in the application economy.
Share