Private cloud or virtualization automation?

Private cloud or virtualization automation?

The private cloud trend started around six years ago when virtualization deployments were growing rapidly and maturing. During such time, virtualization automation became the goal in larger deployments for many organizations.

The majority of private cloud projects have been experiencing problems because of the transformational nature of cloud computing on both the provider and the user. Hence, most organizations have shifted their focus to virtualization automation rather than on-premise private cloud, sometimes not by name but in functionality.

The Day Zero problems

Transformation of operational models, organizational structures and roles, IT and IT user culture, standards and technology architecture is required in on-premise cloud computing. Most private clouds are on-premise and based on existing virtualized environments. At a poll taken at Gartner Data Center Conference, 95% of the respondents had problems with their private cloud deployments. These problems included failures to transform culture, process, organizations or business models or that the private cloud was focused on delivering the wrong kind of services.

Less cloud and more virtual automation

?Most of the on-premise private clouds are relatively small. The larger deployments tend to be less cloud and more of a virtualization automation, with a little use of self-service and standard offerings, lots of customization, some amount of manual effort (with the exception being a smaller web-scale, leading-edge enterprises. According to a survey, the number of respondents who had some kind of private cloud deployment project in place was relatively unchanged between 2012 and 2014.

The two forms

There has been a strong evidence that there are two types of on-premise private cloud projects. The first and the less common form can be said to be the true cloud form, with standard offerings delivered through automation, typically for development and test environments or built for very specific uses. The other and more common form of private cloud is probably private cloud in name only. It is really infrastructure modernization, or virtual automation which cannot be called private cloud at all. We can look at VMware, which is also used for private cloud deployments (specifically for agility in provisioning, as in development/test), but literally dominates the market for virtualization automation.

Recommendations

A few key recommendations for on-premise cloud users can be:

  • Be aware that agility is key for true private clouds. Position private cloud efforts appropriately, for example in Mode 2 application.
  • Reduce the transformation pain by employing third parties. But also remember there are going to be some transformational challenges in effectively using and governing your use of cloud services.

Understand that private does not necessarily mean on-premise. Hosted services can also be chosen, or even public cloud offerings when your isolation requirements are adequately met.

Private cloud should only be used when there is no alternative left. Public cloud services have been improving rapidly and most of the organizations have discovered that public cloud offerings can meet their requirements. But if the use of private cloud for cloud computing is inevitable, then consider the above mentioned recommendations and keep in mind the kind of challenges you might have to face.

#BringItOn

要查看或添加评论,请登录

社区洞察

其他会员也浏览了