57: Nutanix Weekly: Approval Policies in NCM Self-Service available as Tech Preview

जुलाई 26, 2022

We’re pleased to announce that the Approval Policies feature is now available in Tech Preview for the Nutanix Cloud Manager™ (NCM) Self-Service.

NCM Self-Service (formerly Calm) provides an easy and seamless way to consume simple Infrastructure as a Service VMs and multi-tier applications via the marketplace[JG1] . This is possible via easy to use blueprints to automate complex application deployment processes. As NCM Self-Service continues to grow its footprint across enterprise deployments, customers have remarked on the need for even better governance and controls on resources provisioned and managed. Governance is key to building a cloud-like experience for IT organizations, and the ability to easily define and monitor policies are important for enterprise customers.

होस्ट: हार्वे ग्रीन
सह-मेज़बान: जिराह कॉक्स

Adding Approval Gates to Self-Service Automation

Self-service automation lets teams deploy virtual machines and applications quickly, but speed creates its own risk. Quotas can limit how much a user or team consumes, yet some actions still warrant a second set of eyes. Approval Policies, now available as a Tech Preview in Nutanix Cloud Manager (NCM) Self-Service, add that control.

NCM Self-Service is the product formerly known as Calm, and NCM is the broader product that grew out of Prism Pro and Prism Central. The new feature lets administrators require sign-off before certain automated actions complete.

How Approval Policies Work

Configuration has three parts. First, choose the entity type, such as an application, which can include one or many virtual machines or container launches defined in a blueprint. Second, choose the action that should trigger a review, such as launching a blueprint. Third, set the conditions, such as a memory threshold.

One example describes a production deployment that requests more than 10 GB of memory and therefore needs an approval. Other conditions mentioned include the network being used, snapshot actions, CPU core counts, disk sizes, and an estimated application cost. Approval can also apply to runbooks. Approvers can be any group an administrator defines, and a policy can require one approver from a group or responses from all of them. Once the threshold is met, the deployment continues without further action from the requester.

Guardrails for Junior Staff and Rogue IT

The practical value is easy to see for IT managers. A junior engineer can be given access to build environments while a more senior colleague reviews requests that would consume significant capacity. That avoids restricting people outright while still protecting the platform from an overly ambitious request, whether it comes from a junior engineer, a developer, or a database administrator sizing a very large database.

The same logic helps with unplanned cloud spending. Running on-premises infrastructure is already paid for, but building in a public cloud adds cost, so an approval step before that deployment helps avoid surprise bills.

Works Across Clouds

NCM Self-Service blueprints are written once and can be deployed to several environments, including vSphere, AHV, AWS, Azure, and GCP. Approval policies apply across those targets, which makes them useful in a multicloud setup where the same governance should follow the workload.

Built In, Not Bolted On

Before this feature, organizations typically routed approvals through a separate portal and then called into Calm once approval was granted. That works, and platforms built for complex business routing do it well, but it adds integrations to build and maintain. Here, enabling NCM Self-Service takes one step and turning on approval policies takes another, all within Prism Central with no external services.

Approval policies complement existing safeguards such as project quotas on CPU, memory, and storage. Together they give administrators several layers of protection against changes that could affect stability.

Status and Next Steps

Approval Policies are a Tech Preview, which is typically a step before a broader rollout. Customers can try the feature and share feedback with their account team. It is also a reminder that production environments benefit from review steps, because human error remains one of the most common causes of outages.

hi_INHindi