软件定义数据中心 (SDDC)

Software defined data centers afford several benefits to both the data center administrators, but also users. Arguably, traditional data centers will become too expensive to efficiently operate, and will need to adopt the profit improving techniques used in their virtualized counterparts. SDDCs have several benefits:

• Business Agility — At a business level, SDDCs are an evolutionary step beyond traditional data centers. Virtualization increases productivity, and eliminates the need to make hard commitments of resources, instead, every last cycle, and storage media can be maximally utilized. Virtualization also grants exceptional flexibility when it comes to choosing software, as well as hardware components to invest and install. Since abstraction decouples the two, new or updated software or hardware can be swapped in and out as needs dictate.
• Reduced Cost — Productively improvements also reduce cost in the long term. Initially, instituting a SDDC requires upfront investments, however, because transparency and system control allow leaders to track real performance granularly, all costs can be controlled and optimized. 
• Increased Scalability — Respectfully, traditional data centers tend to expand more than they scale. Because traditional data centers cannot divide their resources efficiently, as SDDC can, when they need more of anything, they simply add new machines, expanding the number of devices there are. True scaling understands that load is the operative variable, and that when scaling some loads need more than others. Within a virtualized environment, resources can be allocated from idle pools to scale heavy workloads, instead of adding new machines to the system. This is true scalability.

Software-Defined Data Centers (SDDC) extend traditional data center configurations through a virtualization layer. Virtualization is a software approach to creating IT services separate, or unbound, from designated hardware. To do this a hypervisor is used which is responsible for creating virtual environments and pooling and sharing hardware resources.  Hypervisors can sit on top of an operating system or be installed directly on hardware infrastructure. While resources are partitioned on-demand to each virtual environment, users interact with these environments, known as Virtual Machines (VMs) as they normally would. A single computer running virtualization software can have multiple VMs running, and each VM can potentially run their own operating system, enabling unmatched system flexibility.

SDDCs will become standard practice in the future. Making the transition from traditional data center operations to SDDC operations, though, requires forethought. While every SDDC migration journey is different, the following 3 points should be considered when deciding to undertake the migration.

• Assess the company culture as a starting point — Initially, understanding the benefits of SDDC may overshadow what that means for the company culture. Businesses that lean more toward legacy configurations will have the most adaptation to do, and will likely be the most resistant. Search out for key players that can champion change within the culture, and identify those company practices that will likely meet the most resistance.
• Do not underestimate your time horizon — Migrating from traditional to SDDC should be done in stages to ensure that systems have been successful. For example, start by migrating one system at a time, like migrating MS Exchange servers over to MS Azure. By ensuring that time buffers migration stages, enough time will be left to properly setup and guarantee systems are running smoothly.
• Understand your DevOps team capabilities and limitations — Like assessing the company culture and their ability to adapt to SDDC style operations, your DevOps team must be technically prepared to do so. Do they have the requisite knowledge? Can they easily acquire it? Does there need to be an additional team member?