In the first Q-and-A session of this series with Ganesh Padmanabhan, product marketing manager at Dell virtualization solutions, we discussed on-demand services in the cloud. Now, in this installment we take a closer look at provisioning.
For more information on provisioning in the cloud, tune in to E2 Radio on Thursday, Sept. 15 at 2 p.m. EDT. Host Curtis Franklin and guests from Dell and Intel will delve into thin provisioning in the cloud.
In terms of on-demand self-service in the private cloud, is it completely automated? If not fully automated, to what degree will it require human intervention on one side or the other? In what sort of scenarios could a client expect to engage human support?
Padmanabhan: It depends. If the IT consumer requests a service that’s published in the IT catalog without any modifications and the provisioning system is already automated to provide that service, this can be done without any human intervention.
If there is a need for human intervention to ensure governance—such as approvals or to take actions like updating an inventory management system— the private cloud implementation should be able to provide this flexibility.
An ideal method in an on-demand private cloud is to provide support for multiple provisioning automation methods from scripted operating system and application installs to thin cloning to service orchestration, which will provide the flexibility to automate this provisioning method at required levels.
Will service level agreements (SLAs) be extended to include provisioning times?
Padmanabhan: Any impact to SLAs on the negative side may not justify the use of a private cloud. Depending on the current definitions of the SLAs, it may need to be modified to reflect faster provisioning times.
To what extent will provisioning allow customer to easily ratchet up or down, based on demand, requirements, and budget?
Padmanabhan: This is a good question. Using a utility-based model is not enough to have a successful private cloud implementation.
Providing a complete service lifecycle management framework is critical to track the utilization, budget of a service from provisioning to decommissioning, and IT governance framework. This ensures that IT administrators have efficient use of the correct resources for the right people.
In addition, the need for dynamic capacity management based on demand is a key requirement for implementing a successful private cloud along with ensuring elastic availability of resources based on demand and budget.
The Dell advanced infrastructure manager (AIM), a component of the virtual integrated system (VIS) portfolio, is developed with this design point to provide elasticity and workload mobility across multi-vendor hardware, multiple virtualization technologies and across the physical and virtual paradigms. AIM provides the kind of workload mobility that’s unparalleled in the industry and forms an infrastructure abstraction platform for the private cloud.