It’s easy to deploy workloads into the cloud. So easy, in fact, that some organizations are starting to put the brakes on the practice.
In the early cloud days, a unit manager could spin up a few resources on the cloud, run a project, and then decommission the environment when the work was done. It was faster, cheaper and just plain easier than working through IT. But now that the cloud is emerging as a permanent fixture of enterprise infrastructure, the need to devise a coordinated cloud strategy is on the rise. After all, if silo-based infrastructure is inhibiting progress in the data center, imagine what it will be like in the cloud.
A recent survey of nearly 1,100 IT professionals by Evolve IP showed that while more than 90 percent have deployed at least one service in the cloud, and that the average number of services has jumped from 2.7 in 2014 to 4.1 in 2015, the desire to manage growth in the cloud is emerging as a key trend.
The report says that while more than half are deploying cloud services on their own, an 11-point gain over the previous year, a good half of those respondents say they will seek help from a third party, such as a service provider, channel partner or consulting firm, for their next attempt. The results indicate that while self-service is entirely possible in the cloud, it does not necessarily result in optimal service over the long term, particularly when it comes to coordinating applications and services to suit broader business objectives.
This could emerge as a crucial stumbling block as organizations seek to build integrated hybrid clouds. As automated, app-driven service starts to take hold, the underpinnings of the cloud infrastructure have to be rock solid, otherwise the environment might not behave as desired. Platform providers like Cisco are acutely aware of this fact and are starting to integrate key elements of the cloud stack to provide a cohesive hybrid environment. The company recently announced plans to package its Nexus switch portfolio with the HyperFlex UCS platform and the possibly soon-to-be-acquired CliQr CloudCenter application orchestration stack to create an easily deployable hybrid cloud ecosystem.
The idea is to provide plug-and-play set-up of compute and network infrastructure and then layer it with advanced fabric management, flexible bandwidth support and broad automation. The resulting cloud should then be easy to manage, highly functional and occupy a minimum resource footprint.
No matter which platform is used, however, enterprise executives should be aware of the three critical mistakes that usually crop up in the deployment phase, says IT consultant David Linthicum. The first is a lack of governance at both the resource and service levels, which leaves managers unable to put the brakes on initiatives that fail to meet objectives or actually cause harm to the overall environment. Secondly, failure to implement a single management stack makes it possible to recreate the same silos that hamper data center operations. And while most organizations pay strong heed to security in the cloud, it is usually defined improperly so that key pieces like identity and access management are overlooked.
Deployment issues are also one of the key drawbacks to clouds based on open platforms, which still require a fair amount of integration and coordination from highly trained technicians. This is why companies like Red Hat and Rackspace are implementing new OpenStack as a Service models that allow clients to mount a fully managed cloud using a streamlined deployment process. The package is built around the Red Hat Enterprise Linux OpenStack Platform, which can be hosted on-premises, in the Rackspace cloud or both. In this way, the enterprise can manage its cloud as a service rather than a distribution and thus avoid the steep learning curve that usually accompanies OpenStack deployments.
Few people would argue that issues like deployment, migration and management are inconsequential in the cloud. Users want as pleasant and productive an experience as possible, and providers have a strong financial incentive to deliver it. Virtually every cloud will need its own set of tweaks and customizations to suit the needs of individual enterprises, but as long as the basic deployment process can be nailed down, the fine-tuning should become easier.
With a simpler cloud, everyone wins.
Arthur Cole writes about infrastructure for IT Business Edge. Cole has been covering the high-tech media and computing industries for more than 20 years, having served as editor of TV Technology, Video Technology News, Internet News and Multimedia Weekly. His contributions have appeared in Communications Today and Enterprise Networking Planet and as web content for numerous high-tech clients like TwinStrata and Carpathia. Follow Art on Twitter @acole602.