Resource policies give platform operators the ability to use meta-data to publish CPU, memory, and storage plans that developers in the enterprise can associate with their applications. Apprenda allocates resources based on what developers choose, dividing the infrastructure based on these policies.
The Apprenda PaaS stitches together any number of Windows and Linux OS instances (physical or virtual) into a single, logical hosting fabric. Rather than relying on virtual machines as a container for providing application isolation, Apprenda uses a custom container model that runs inside OS instances to isolate applications from one another. This fine-grained isolation is used to sub-divide each OS instance into multiple segments, increasing utilization and reducing hardware and licensing costs. Resource policies are the defining unit that allow platform operators to define how large these segments are from a CPU, memory and storage perspective, and are published to developers as the set of options they may choose from when allocating resources to their applications.
Key Points
- Process and database level isolation increases infrastructure utilization and slashes licensing costs
- Define quotas for CPU, memory, and storage. Drag and drop and manifest based assignment of quotas for developers.
- Publishing system ensures platform operators can control resource options for developers, while allowing developers the freedom to make application specific
resource decisions - Utilize tier specific definitions (frontend, web-services/daemons, or databases) or
publish broad based definitions - Cost definitions for chargeback/showback
- Oversubscription allows for higher policy packing density
- Hypervisorless implementation allows for bare-metal deployments
- View and manage live application components through resource policy filters
Through resource policies, platform operators in the IT department get ultimate control over resource utilization while giving developers the authority to make sanctioned, application level resource assignment decisions. The end result is lowered IT costs through increased utilization and compliance.
