Net Management and the Cloud

Research indicates network operators playing important role

EMA has just recently completed the data gathering phase of our Responsible Cloud research initiative, and there are some interesting results around the roles being played by network management and networking teams in terms of cloud management and operations. While much of the focus on the cloud is on the application platforms, virtualized systems, and storage, accessible remotely as a packaged service, there are also big questions about the role that networking will play. All of those services must be accessed via the network, whether they are internal resources wrapped in private/internal cloud architectures or they are external resources residing in the public cloud. In fact, when there are external services that will be utilized, those services will have to be reached over the wide area, which brings with it the specter of significantly higher latencies as well as potential bandwidth constraints. Further, how do you visualize the managed environment, when some of your infrastructure is yours, and some of it isn’t? Where are the hand-offs? Where are the connection points? And how do you troubleshoot the inevitable calls that the network team will get when responsiveness is not meeting expectations? These would seem to be some serious concerns, however only 18% of our respondents indicated that they had either experienced or expected to experience poor service quality in terms of more downtime, slow response, or missed SLAs when moving to clouds. Curiously, at odds with this lack of concern were later responses indicating that Performance and Availability is the most important service monitoring discipline – the only discipline characterized by more than 50% of respondents as being “very important.” Another fascinating discovery was that Network Operations is emerging as the most likely functional area of the organization which is or will be responsible for cloud service management – more than security, cloud/virtualization support, or data center/server operations! This might seem an expected response if the audience was largely a networking audience, but our research was gathered from a broad mix of practitioners, only 6% of whom claim to be part of the Network Operations group. When it comes to metrics that are in use or planned to be used, network-based measures also ranked very high. Overall application response time, which can be measured via passive network monitoring, was seen as second most popular, and network infrastructure performance ranked third. The only metrics that ranked higher (and only slightly so) were system or application availability (i.e. uptime). Granted that application response time can also be measured using other technologies, however if application-aware network infrastructure performance monitoring tools such as packet-based inspection are in place, then both can be derived from common, shared instrumentation. Further underscoring the importance of the network, our respondents indicated that network-specific monitoring tools are the most important type which they have integrated or are planning to integrate into their cloud computing monitoring systems – more important than tools specific to applications, systems, security, or storage. It’s worth pointing out that all of these categories received at least 50% acknowledgement and that network-specific was tops at 62%. And it’s also worth noting that BSM/SLM monitoring for cloud was a distant trailing priority, garnering a mere 27%. Expect much more detail to come as EMA completes analysis of this initial round of research into cloud management and operations. But for now, based on this initial pass through the results, it is clear that network monitoring and the network viewpoint is playing an important role in the mind of practitioners across the spectrum.

Copyright © 2009 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022