CycleCloud Overview

Key Features

Using CycleCloud

 

Cycle Computing’s CycleCloud software suite is the leading cloud orchestration, provisioning, and data management platform for Big Compute, Big Data, and large technical computing applications running on any public, private, or internal environment.

For years, customers in Life Sciences, Manufacturing, Financial Services, and other Engineering and Research areas have used CycleCloud software to manage some of the world’s largest production cloud deployments. The ever-increasing demand for support of compute and data intensive workloads, machine learning, genomics, data collection, analytics, and simulation is causing many organizations to struggle to meet the needs of their users.

Cycle Graphic

Cloud based workflows deliver:

  • zero queue times
  • unlimited amounts of a wide variety of architectures
  • ability to fit jobs to budget/value instead of available capacity
  • dynamically adjustable infrastructure to service a wider range of applications
  • capacity to match spending to actual consumption

While it is easy to get a cloud server up and running, it is only productive once a full cloud workflow is created. Overall management, control, and efficiency of the workflow can be a challenge.

  • Users have the ability to exploit unlimited compute capacity, but often need assistance to exploit it effectively.
  • Optimizing and managing multiple cloud providers, especially given the various platforms they each provide, is complicated.
  • Placing workloads in public clouds raises security concerns that need to be addressed.

Running multiple workflows without tools can quickly become overwhelming. Leveraging the CycleCloud tool suite solves these issues plus more for users, system administrators and management, while delivering the maximum value out of any cloud workflow.

Key Features

CycleCloud has been engineered for enterprise-class production deployments and manages secure, large-scale, and resilient cloud environments for some of the world’s largest organizations. With over one-hundred engineering-years of development, CycleCloud is:

  • Resilient – Leveraging template-based, provider-agnostic infrastructure definitions, CycleCloud can provision and synchronize a copy of your infrastructure in another provider data center, or even another cloud provider. This capability is unique, and patent-pending and simplifies disaster recovery planning while providing a provider-neutral option to your organization.
  • Secure – CycleCloud conforms to industry best practices at every level of the stack. It leverages cloud provider-specific security mechanisms to control network, compute, and storage access in a repeatable, controlled manner to help ensure consistency in deployments. CycleCloud can directly integrate with on-premise authentication and authorization systems such as LDAP and Active Directory, and offers API-level integration for a host of other configurations.
  • Flexible – CycleCloud’s configuration management tool provisions the entire software stack including leading operating systems, HPC, HTC, and storage configurations out-of-the-box. It is also fully customizable allowing you to easily incorporate custom configurations with bring-your-own machine images and Chef cookbooks.
  • Scalable – CycleCloud can handle any scale, from tens of servers to thousands.

The CycleCloud orchestration suite manages the provisioning of cloud infrastructure, orchestration of workflow execution and job queue management, automated and efficient data placement, full process monitoring and logging, all within a fully secure process flow. As shown in the diagram above, CycleCloud easily leverages multi-cloud environments moving seamlessly between internal clusters, Amazon Web Services, Google Cloud Platform, Microsoft Azure and other cloud environments.

The solution provides a web-based GUI, a command line interface, and a set of APIs to define cloud-based clusters. Once defined according to policies set by system administrators, CycleCloud can auto-scale clusters by instance types, maximum cluster size, and costing parameters. It rapidly deploys everything from modest sized systems of 64-6,400 cores to systems that rank as some of the fastest computers in the world (156,000+ cores), while validating each piece of the infrastructure to insure a complete and robust environment.

CycleCloud capabilities include:

  • Provision, manage, and orchestrate cloud infrastructure from multiple providers
  • Full-featured web GUI , command line tools, and APIs
  • Purpose-built, cloud-optimized, configuration management system
  • Dynamic scaling of large computation, Big Data, Big Compute and HPC workloads
  • Utilization reporting, logging, and auditing capabilities
  • Job submission, monitoring and administration of cloud deployments

CycleCloud can automatically scale to meet the needs of applications by elastically provisioning infrastructure based on user-defined templates. These templates offer complete control of the operational characteristics allowing dynamic optimization of workloads over multiple dimensions such as cost, scale, resiliency, and performance. Additionally, the data management function syncs in-house data repositories with cloud locations in a policy / job driven fashion. This enables data driven batch submissions where compute infrastructure is only provisioned when data is in place, saving costs and improving efficiency.

 

Cycle Workflow Graphic

CycleCloud is designed as a complete, multi-cloud workflow solution. As shown above, the CycleCloud solution manages all the key elements for a full cloud based workflow. These include cluster configuration, provisioning, monitoring, and optimization. CycleCloud handles all of these in a programmatic fashion and insures that any options selected can be leveraged across the full range of providers. Keeping track of the full workflow, leveraging the feedback loop through monitoring, and providing this across multiple workloads simultaneously makes it possible to rapidly expand computational capabilities without major modifications or delay.

Using CycleCloud

Step 1:

Cycle LifeCycle 1

To provision a compute environment, the user begins by selecting a cluster template that specifies the composition of the infrastructure (instance type, storage and network configuration), the software configuration (role-based service deployment), and the dynamic behavior of the environment in the cloud (provisioning behaviors, optimizations, resiliency, cost profiles). Using this template, CycleCloud uses the cloud provider API to instantiate the infrastructure, configure the software stack, and prepare for use.

Step 2:

Cycle LifeCycle 2

Users access the cloud-based cluster over secure applications such as ssh or RDP over a VPN connection. The initial state of the cluster is determined by the cluster template selected in Step 1 and typically consists of a master node, a shared filer/storage and any unique configurations required by the specific workload. CycleCloud monitors the cluster job queue and, based on auto-scaling parameters in the template, dynamically provisions fully configured execute nodes.

Additionally, depending on the template rules, CycleCloud can automatically bid for resources at different price to take advantage of potential cost savings associated with spot markets. This can significantly lower computational costs.

Step 3:

Cycle LifeCycle 3

As the user’s jobs complete, CycleCloud verifies the job completion, manages final data placement and tears down cluster resources on billing boundaries to help control costs. This behavior is configurable and helps ensure that you only pay for what you need, when you need it.

With today’s complicated workflows, increasing system demands, and rapidly changing marketplace for cloud offerings, having an orchestration suite that is designed from the ground up as a single control point is critical. Cycle Computing’s CycleCloud software suite delivers this ability with its leading cloud orchestration, provisioning, and data management platform for Big Compute, Big Data, and large technical computing applications running on any public, private, or internal environment.

View our documentation for additional technical background

 

 

ShareShare on FacebookShare on Google+Share on LinkedInTweet about this on TwitterEmail this to someone