Elastic Computing – Dynamically Scale Clusters
While some organizations have the benefit of dealing with consistent workloads, most experience bursty workloads. These workloads vary in quantity and size, which makes capacity planning an exponentially more difficult task. During peak times, administrators must ensure that regularly scheduled workloads are completed and associated SLA’s are met. Large system administrators juggle lots of users’ needs and the requirement to be responsive to those needs is imperative; therefore, being able to burst workloads to other resources becomes extremely desirable.
Elastic Computing gives HPC administrators the ability to manage resource expansion and contraction to increase productivity by providing additional resources when required. By bursting to private clouds or other data center resources utilizing OpenStack or other standard platforms, elastic computing helps admins better manage the provisioning and performance challenges of bursty workloads.
- Dynamically scale your cluster with additional capacity
- Burst to private or public clouds, including such resources as AWS, Azure, and other Cloud API’s (e.g. Amazon)
- Better manage provisioning and performance challenges during peak workload times
Automation with Moab
As workloads increase, Moab automates the growth requirements and dynamically obtains additional resources from other data center resources to handle the peak loads and then relinquishes those resources back to the original resource for the next peak workload requirement.
Moab increases productivity with elastic computing, which allows admins to efficiently manage resource expansion. Elastic computing is triggered when a threshold set in Moab is exceeded. To determine this threshold, Moab surveys the system workload and calculates the combined completion time of these burstable workloads if no other workloads are running. Elastic computing bursts workloads, on an as-needed basis, into a communal pool of data center resources and then relinquishing these resources back to the shared pool. Using tools like OpenStack, CMU, Bright Cluster Manager, Moab completely wipes each resource after use to help comply with privacy regulations. This added flexibility enables admins to expand their own cluster while taking advantage of the elasticity of resources and scalability of the cloud.
Below are animated illustrations of two principle use cases. The first enables a local system to add or remove resources from another local resource pool. The second dynamically carves out resources from a single shared resource pool in a secure way to facilitate multi-tenant requirements. The second use case also dynamically load-balances resources between the tenants to maximize efficient utilization.