Self-service provisioning is good for everyone in the food-chain. It allows customers to do the key day-to-day administration themselves, and removes the need to raise support tickets, which makes them happier and more loyal. It also allows IT to spend more valuable time looking for new business challenges that the customer may face, rather than endlessly fulfilling new requests or doing day-to-day mundane and repetitive tasks.
What to Look for in a Self-Service Portal
A good self-service portal presents the right information in the right format and requires little support, so users can be self-sufficient. In making it easier for users to provision their own machines, the environment can increase in scale, and, if not correctly implemented, can bring with it VM management challenges. These include:
- difficulty keeping track of the systems and their owners
- reasons behind VM creation such as environment flexibility and consistency being lost
To counter these possible downsides, IT needs to maintain a high-level of awareness over how resources in the environment are being used, and the impact across all areas, not only hardware, but also costs and administrative overhead.
Making Self-Service Provisioning Economically Viable
If self-service provisioning is to be economically viable, you need to use available resources as efficiently as possible, and the first area to look at is the scalability requirements within your infrastructure. In a self-service environment, IT is unable to restrict consumers to a point where they don’t have enough resources to do their jobs. Doing so would open up the possibility that users would just bypass IT and start to use public cloud resources without either the knowledge or consent of IT, a situation often referred to as “Shadow IT”. If shadow IT creeps into an organization, requirements for control, documentation, security, and reliability can soon be dropped by the wayside in a bid for increased business agility and productivity.
This is where scalability needs to be taken into account. The nature of the request needs to be considered, and IT should be able to implement a solution that is well-suited to the situation. For example, if there has been a recent influx of new employees to a particular department, then adding additional servers to accommodate their needs might be justified. If on the other hand, additional resources are required for a project that will be completed in a few months, then it may be better to temporarily leverage the public cloud. Either way, it’s better for IT to be a part of the conversation, and to enhance the offerings within the self-service portal with the best possible options, than to be sidelined and have users operating outside of the portal where you can’t track usage.
Reducing VM Sprawl
So, now you have the scalability issue sorted, what you don’t need are unmanaged virtual machines sitting out there that can negatively impact the bottom line. Wasting resources through virtual machine sprawl has an associated opportunity cost because those resources are not available for use by other virtual machines.
Reducing waste and virtual machine sprawl sounds simple enough, but it can be deceptively difficult, and needs attacking on two fronts. First, measures must be taken to stop virtual machine sprawl from continuing in the future, and second, and more challenging, is to go back and address any previously existing sprawl or waste.
Implementing an automated lifecycle management system, assigning default expiry dates on the initial request is a good first step to preventing sprawl in the future, along with the implementation of a chargeback system, as people are far less likely to waste resources when they are being billed for those resources. But even if the virtual machines are tied to an automated process, the system administrators still need insight into the process, and running a reclamation report will allow administrators to see which resources have been reclaimed and which virtual machines have been removed.
Other reports allow administrators to locate virtual hard disks that are not currently associated with a virtual machine and rightsizing reports can analyze performance monitoring data across the infrastructure and recommend changes to resource allocations. By regularly running these reports and reclaiming wasted resources, it’s possible to achieve a higher overall virtual machine density, which, in turn leads to a reduction in costs.
Effectively Managing Users
The final area to look at to limit cloud challenges is that of the users themselves. In the same way that you don’t want to unnecessarily restrict the scalability options, you also don’t want to implement a strict set of usage policies and permission models that are designed to prevent customers from getting access to the services that they need These policies need to be implemented to act as a protective mechanism for the underlying infrastructure, so that a user consuming an excessive amount of system resources causes resource contentions that may negatively impact other consumers.
Creating a good permissions model is one half of the story, as the creation of the virtual systems also need to conform to the governance rules of operation that your organization has established. In most cases, for example, it won’t be appropriate for users to create virtual machines at will, as this will impact the department’s budget. It’s far more effective in situations like this to allow consumers to request a virtual machine, but then have it approved by a manager who oversees the department’s resources and budget.
Alternative Self-Service Provisioning Models
Another method of allowing users to self-provision, but still providing checks and balances, is to employ a quota system for the resources based on their business requirements. Self-service provisioning environments generally use quotas for two primary purposes: first, to track the cumulative resources that a user may consume, and second, to limit the consumption within a department. This enables IT to set a resource consumption cap for all of the user’s and department’s virtual machines, and, as long as requesters stay within their quota limits, removes the need for an approval process. If there is a temporary need for additional resources, an approval workflow can be put in place to ensure that any request over-and-above the quota can be approved, but it is done with full awareness of the impact and how it could affect other projects.
Quotas can be based either on resources (vCPUs, RAM and storage) or on a cost basis, which, along with chargeback/showback reports, gives the cost and budget visibility that departments need. This also provides the IT department with a means to monitor and prevent, the depletion of available hardware resources.
Self-service not only speeds up service request fulfillment, but when done correctly, will bring a reduction in costs and a more hands-on customer experience, leading to a feeling of more independence. Other potential benefits that can also be realized through automation of the cloud infrastructure are:
- Better workforce utilization. Automation is about reducing the reliance on manual labor – especially for repetitive tasks, tasks that people don’t necessarily like to do, and tasks where the automation can not only do things more quickly but with a better result. Self-service also frees up highly-skilled and knowledgeable staff from repeatable, often mundane, tasks to focus on more “intelligent” work.
- A reduction in human error. The benefit here is not so much avoiding the error itself but the unwanted consequences, such as the business impact of the error. As the Principal Data Center engineer at a leading CRM software company recently said in a case study, “In addition to the time saved in provisioning, we also reduced human error. It wasn't a calculated thing, but the reduction there was huge.”
- Greater ability to change. Automation allows processes and procedures to rapidly, and consistently, adapt as both business and IT requirements evolve.
User satisfaction and a rise in demand for services are signs that your cloud initiative is valuable and efficient, and demonstrating these improvements helps to justify the time and money spent on implementing automation, orchestration, and governance alongside self-service provisioning.
Enabling Self-Service Provisioning with a Cloud Management Platform
A Cloud Management Platform (CMP) such as Embotics vCommander provides all of the automation and management you need to ensure that you can grow your virtual environments effectively in a single, fully integrated system, without the lengthy and expensive customization and implementation projects that traditional automation systems need.
The ROI of vCommander can normally be found simply from freeing up resources that have been identified through its automated sprawl detection alone, leading to other rewards such as governance, consistency, reduction in errors and time savings being achieved for free. This immediate return gives you the business agility to drive towards greater profitability, faster time to market and the focus and resources needed to create new and innovative products.
As you move towards adopting a self-service cloud computing model and find yourself trying to quantify the return-on-investment (ROI) of your cloud infrastructure, download this whitepaper to see how easy it can be to implement automation into your environment, and how to understand the business case to do so.