Johannesburg, 29 Nov 2017
Cloud is where it's at. The benefits are being shown time and again as more companies adopt cloud into their technology environments. Gartner estimates that already more than 80% of x86 workloads are virtualised, many of which do so in cloud environments.
These shifts are happening in a hybrid world, where workloads and applications are found both inside and outside a company's own infrastructure. Healthy amounts of workloads are appearing on both private and public cloud systems, driven by cloud economics. A study by network vendor Cisco found that cost, resiliency, scalability and product lifespan are the leading reasons making hybrid cloud attractive.
"The benefits of hybrid cloud are fantastic," says Andrew Cole, MD of Concilium Technologies. "But with those benefits come distinct and large challenges. Cloud can enable several workloads distributed across numerous servers to serve a single end-user application. This creates a very dynamic networking environment with multiple streams of traffic. If you can't keep on top of that, you will soon have big problems."
Managing that torrent is complicated by user expectations of exceptional uptime and peak performance, made trickier by the very different journeys different applications follow. Most companies already operate their e-mail and Web presences on outsourced systems, a route followed today by analytics applications, mobile platforms, digital marketing content and CRMs. Yet legacy and core services such as ERP and financial applications are making new homes on private cloud infrastructure.
Hybrid infrastructure is complex. A modern business technology environment is a conglomeration of networks, applications, virtual servers and storage appearing on both the private and public server domains. Yet staying on top of that is not enough, not if the business wants to realise the value and savings it anticipates from its cloud adoption.
"If you don't know what is going on, you won't know what to do next," says Cole. "That's obvious advice, but I rarely see it followed when companies adopt complex hybrid environments. They are gung-ho about getting the new systems in place, but then neglect performance transparency and control. They expect their old monitoring systems to be up to the task, even though detecting and managing relationships between the various parts are more crucial than ever."
Planning is key to success with cloud, and effective monitoring is key to planning. This is particularly important for network monitoring, as the network brings the many components together and also often dictates performance. Yet often companies rely on monitoring systems that were designed for monolithic solutions, cobbled together under one vendor's roof.
This is a far cry from modern hybrid environments. Though it is possible to augment legacy monitoring tools for new roles, the ideal is to invest in a new generation of monitoring capacity. One would want to do this for several reasons, Cole concludes:
"You want to be able to monitor core systems, as well as provide for multitenancy. Silos will always exist, but a multitenant approach makes it easier to monitor and report across those silos. For the same reason the monitoring solution should consolidate events in a single pane of glass and enable administrators to drill down where they need to. Other considerations should be the system's ability to scale as the environment does and that it is flexible enough to accept future standards of performance data as vendors introduce new technologies. Another perk is that some of the new monitor platforms can be deployed in a matter of hours."
Hybrid cloud is the future of modern business. But what looks like a miracle of simplicity to the end-user belies a delicate and complex balance in the background. The ability to monitor that complexity is the cornerstone of getting the most performance and cost out of cloud. After all, you don't know what you don't know - and if you skimp on effective monitoring of your cloud network, you really won't know much at all!
Share