The key to implementing complex projects is to break them down into smaller, less intimidating steps – ones that can be easily focused on and dealt with one at a time. So, when it comes to embarking on a cloud management project, the same rules apply, and there are a common set of management and automation features that are required. The good news is you don’t need to tackle them all at once, but you can take small steps with each to ensure success and adoption. These include:
- Resource Optimization
- Lifecycle Management
- Workflow and Automation
- IT Costing and Chargeback
- Self-Service Provisioning and Service Catalogs
Let’s take a look at each of these features in turn.
One of the fundamental objectives of any data center, whether physical or virtual, is to ensure that you’re configuring resources properly and that you’re using these resources efficiently and at optimal levels.
This means monitoring and managing the configurations and tracking any changes in the environment. You also need to proactively plan future capacity requirements for the infrastructure. The best way to do this is by understanding trends in growth and performance and by identifying VM sprawl and optimizing over- or under-provisioned assets through rightsizing and intelligent workload placement. By using automation, you can send regular sprawl reports to responsible individuals for review and/or action, quickly and easily eliminating sprawl from your environment.
The lifespan of virtual machines can be anything from minutes to years. This variable lifespan requires lifecycle management designed for dynamic IT services. With proper management, only approved and compliant VMs are provisioned, they are managed and controlled while they are in the environment, and they are decommissioned at the end of their life in order to free up valuable resources for reuse, rather than allowing them to languish and run continuously. All of the tasks involved in the VM lifecycle are discrete, well understood, tedious to perform manually, and require a high level of standardization and governance. They’re all ideal first steps for automation.
By leveraging a Cloud Management Platform (CMP), you can easily introduce and monitor the lifecycle decommissioning process. You can also target a specific area of the organization and easily target only those areas of public cloud consumption where VMs may be used for short-term use such as R&D or QA. Automated alerts warn system administrators and VM owners that the established expiry date approaches. This provides the option for owners to extend the VM’s life with a single click in an email.
Workflow and Automation
As server virtualization continues to mature within an organization, and the number of virtualized machines increases, manual environment maintenance and support becomes unreliable or unworkable. IT will start to find that many of the day-to-day management tasks are repetitive, time consuming, or prone to human error. By implementing automated workflows in conjunction with policy-driven automation, you can effectively deal with a large, complex infrastructure spread across multiple clouds. This will also give you the means to enforce your organization’s governance policies without increasing staff.
Our experience with customers has shown that, like implementing an overall cloud management project, the best approach to implementing automation is to implement it gradually. Initially focus on areas that are a clear fit for automation and that will also provide the best overall benefit. This leads to more immediate and quantifiable value, but also lets you work with your consumers to determine the services that should be in your catalog and how consumers will be using the cloud for their workload placement. This also provides value for DevOps teams who need to test early and often, so that errors are more easily diagnosed and repaired.
Automation is critical so manual errors don’t creep in. Automatically deploying requested VMs with the latest configurations enables the automation of unit / system / production testing and intelligent workload and application placement, and avoids those manual errors being introduced.
IT Costing and Chargeback
Cloud architectures are based on the concept of “on-demand computing,” which can be dangerous to implement without the corresponding cost and consumption monitoring that ensures the benefits of elasticity are not left unchecked. In addition to associating costs with specific hardware and software items, it’s important to compare costs within your infrastructure, such as production systems, development systems, systems with high-end or low-end compute/storage resources, or even across different groups of your consumers.
You can control resource consumption by making users accountable for the resources they use and by designing the cost model to change their behavior (charging more for memory makes them think twice before asking for 1TB of RAM!). This method curbs resource consumption and allows users to request resources, but also makes them accountable for the usage. For this method, you can use chargeback, or if the organization is not ready for cross-charging departments, show-back, so that users can see the cost implications of their request, and in parallel, quotas can be assigned to enforce the consumption model.
The second method to control consumption is one of cost recovery. This is where you charge as close to the actual costs as you are able to. This again makes users accountable for resources they consume and means IT costs can remain neutral, spreading the costs fairly across the company.
Once achieved, chargeback places IT in a strategic position that builds financial transparency through reports providing cost visibility, cost comparison and chargeback, which can be used to educate business units on correct workload placement based on cost and technical requirements. It also creates a business-value focused environment, ensuring that you’re focusing resources on demand-driven areas, and increasing confidence in setting and managing IT budgets.
Self-Service Provisioning and Service Catalogs
As your environment grows, Self-Service Provisioning and Service Catalogs become key features that offload administrative work from infrastructure management teams and empower business teams by providing them with ability to request IT services on demand and to track the request all the way to VM provisioning.
Not only are users allowed to self-provision VMs, they are also able to run real-time reports on “their” environments without involving an administrator. You can also delegate other actions down to the users. This includes basic controls such as allowing users to reboot their VMs or providing console access to their systems, without the need for an infrastructure management account on VMware vCenter or AWS. All of the actions taken by the users are tracked and logged, providing an audit trail of who has done what in the environment, to meet governance requirements.
A Collection of Tools or an Integrated Management System?
You will find that at different stages of your cloud adoption journey, different management needs can and will surface, and it may seem easy to satisfy these requirements through the purchase of individual “tools.”
But if you step back, you’ll realize that there’s a great deal of interaction between each of the features listed above. For example, discovery and reporting will feed into every other feature, when implementing self-service, that will feed into the lifecycle management of the systems as well as change and configuration management, which in turn feeds into resource optimization, and so it goes on.
Each one relates to, and relies to some degree, on each of the others, requiring the level of integration that you can only get from a fully integrated management system and one that can manage, and provide insight into, the multi-cloud infrastructure that you will inevitably run your services across.
So, if you would like to see a customized demonstration of the advantages of an integrated management platform, along with the best practices for these disciplines that are included out-of-the-box, click on the link below.