The frenzied excitement around cloud as the "next big thing" in the IT industry is beginning to subside as businesses move from testing the waters, with a couple of applications and use cases, to moving large-scale processing and mission-critical jobs to the cloud. As companies explore cloud adoption models for a range of workloads — IT infrastructure, business applications, and corporate communications — basic service-level guarantees will not be enough. Moreover, as communications service providers (CSPs) expand their offerings to include cloud infrastructure solutions, they must also ensure the quality of the network supporting cloud enabled services. However, the network component need not be merely "dumb pipe" connectivity. CSPs can add application-awareness to network transport with the same types of optimization and visualization technologies being employed by IT/network managers in enterprise WANs.
In corporate networking environments, all fingers point to the connectivity component when applications don't perform as expected. To target the root cause of performance problems, transparent visibility into IT and the application layers is necessary. In cloud computing/IaaS environments, the problem is magnified because resources may reside in different locations, subject to the control/responsibility of different entities. Furthermore, there are multiple moving parts — the cloud server/storage infrastructure, the applications running on the infrastructure, the interactions and interdependencies among the applications and infrastructure, the application resource usage profiles, and the manner in which end users access the applications.
A key benefit of cloud computing is the dynamic resource consumption approach that it enables. However, the dynamic nature of cloud also makes it difficult for service providers and their customers to have unified visibility. Today's increasingly hybrid corporate networks require "smart" pipes that link enterprise infrastructure and applications wherever they reside: on-premise or in third-party service provider datacenters. This holistic perspective can be enabled through application-level visibility and service visualization.
Enterprises cite security and loss-of-control concerns as the primary adoption inhibitors for any type of outsourced IT/network service. These factors become exacerbated when service provider cloud environments become part of the discussion. However, these concerns can be alleviated by providing customers with real time application-level visibility.
Visibility tools allow service providers to offer customers insight into the operations of their cloud deployments; offering insight into key metrics such as resource availability, latency, throughput and utilization. But application visibility takes things a step further by mapping the cloud infrastructure resource components and the associated operational performance to network traffic flows to get a baseline view of how cloud applications perform in context (i.e., how the network is used to deliver the cloud service to the end user).
This comprehensive approach of cloud service visualization offers numerous benefits to the cloud customer and value-added service opportunities for CSPs. By incorporating network connectivity into the cloud service management mix, customers that purchase cloud computing from their Network Service Provider will also receive an end-to-end view of how their cloud-based services and utilities perform across the network, between CSP datacenters and the enterprise datacenters.
Furthermore, CSP-enabled monitoring, management, and visibility can provide cloud customers with valuable intelligence on ways to optimize their infrastructure and network deployments for cost, scale, and performance. For example, service visualization could reveal that a customer baseline usage of cloud IaaS (as shown by network flows) warrants additional infrastructure capacity or a different consumption model (pay-per-use or the ability to burst above capacity for peak loads). For the CSP, the intelligence gleaned from translating packet-level intelligence into application-/process-centric metrics can offer advance notice of potential up-selling/cross-selling opportunities. Furthermore, this level of visibility can be used to benchmark normal/acceptable network performance and to determine thresholds for alerts or triggers for dynamic resource allocation. Comprehensive monitoring capabilities and portal-based service visualization provide CSPs with an important source of service differentiation — elevating cloud-based IaaS from a commodity-oriented offering to a value-added, application-aware IaaS service.
This post is a part of our continuing Cloud Services Assurance Expert Blog Series featuring contributions from the analysts of IDC.
View part 1: Communications Service Providers (CSPs): Claiming a Seat at the “Cloud” Table
by Melanie Posey, Research Vice President, IDC Hosting Infrastructure and Managed Network Services.
View part 2: For the CSP Cloud Customer: Seeing is Believing
by Chris Cullan, Product Marketing Manager, Business Services Solutions, Infovista