Moab Utility/Hosting Suite™
Moab Utility/Hosting Suite allows organizations to dynamically generate whole compute environments–including compute, service, node, network, storage and license resources–to meet changing workload levels, cluster health and resource failures. Moab creates a virtualized pool of all resources and then activates disk-full, diskless, and virtual provisioning technologies to adapt the resources based on workload and specific user or group needs. It fully automates resource allocation by matching workload to the optimal set of resources.
Moab also provides advance reservations and ensures that Quality of Service and SLA guarantees are always met for both internal and external customers. With Moab, organizations can offer needed resources, tools or services to external customers in near real time. And Moab does all this transparently with no changes to the user behavior. Moab Utility/Hosting Suite enables organizations to consolidate all of their resources for maximum utilization while still protecting the sovereignty and flexibility of different users and groups.
Uses for Moab's Utility/Hosting Suite:
- Enable environment to dynamically adapt to constantly changing workload needs to maximize utilization and service levels
- Host out tailored resources to other groups–internal or external–to optimize utilization or for cost sharing purposes
- Instantly Utilize commercially hosted resources to overcome workload spikes, cluster health issues or resource failures by transparently accessing tightly integrated, tailor-provisioned hosted resources
Business Benefits
- Dynamically adjust the entire environment to meet workload needs with dynamic provisioning and allocation of the full virtualized pool of resources based on backlog, QoS/SLA guarantees, resource failures, or virtually any desired factor
- Create a custom packages of all resources each group needs with Virtual Private Clusters
- Host out and bill on the range of resources used, including CPU, network, storage, applications, etc.
- Allow jobs to automatically run on hosted resources when the local resources can not meet needs
- Let users transparently access and pay for unique resources as needed to avoid purchasing expensive, under-used resources and deliver optimal service
- Allow local administrators to control the hosted resources as if they were part of the local cluster

Key Capabilities
Dynamically adapt environment to improve utilization
- Harness the full value of existing compute resources with holistic management that enables resources to be virtualized, viewed, managed, allocated and provisioned automatically to meet changing workload demand
- Allow compute resources to be reserved in advance and jobs and data to automatically migrate to the most appropriate resources to meet the needs of internal or external workload
- Accommodate the resource needs of different types of workloads (transactional, batched, interactive or calendared) with the best set of available resources
Share under-utilized resources across groups
- Unify your environment into a virtualized pool of resources to share under-utilized or specialized resources (CPU, storage, applications, etc.) with other departments and groups, with usage billing if desired
- Use detailed resource consumption reports to effectively share costs of cluster maintenance or to drive self-management behaviors across users, groups, and organizations
- Centralize resources for management and utilization efficiency while protecting the flexibility and sovereignty of local clusters
Guarantee improved service levels and reliability
- Access hosted resources to grow and shrink your local resources based upon factors such as workload spikes, backlog, QoS guarantees, performance metrics, resource failures, etc
- Set up your local system to automatically allocate and de-allocate hosted resources based on conditions such as workload, time of day and resource failures or to guarantee resource availability and fulfill existing SLA and QoS based arrangements
- Use policies and triggers based on events, conditions and time to ensure usage matches set levels and priorities
- Ensure no one exceeds intended usage with soft and hard limits for users, groups, projects or organizations
Provide leading resource control to administrators and ease-of-use to end users
- Create triggers based on any event, policy or condition to automate desired administrative tasks or responses
- Reduce end-user job management training, as users simply submit jobs to the local cluster as usual or via an easy-to-use and fully customizable Web interface
- Use script translation to let users continue job submission unchanged, with Moab translating jobs from the local format (TORQUE, LSF, etc.) to the hosted system’s format (LoadLeveler, PBS Pro, etc.)
Gain additional revenues by hosting out under-utilized resources
- Create new revenue streams by hosting access to under-utilized applications, services and resources that customers can either access on an evaluation basis or pay based on consumption
- Send/receive monthly automated billing for hosted resources in a phone-bill-like report that shows consumed resources, tracks ROI and monitors current and historical job and resource trends
Control information sharing and privacy
- Enable dynamic per user or per job networking host security to protect you and your customers
- Use Virtual Private Clusters to let other organizations, groups and users only view resources, workload and policy information that pertains to them, allowing full resource details to be kept private
- Allow all desired resource, workload and policy information to be shared for detailed accounting and reporting
Example Sites Using Utility/Hosting Suite
Hardware Providers
- Let potential customers test their code and partners test their products on your hardware. These customers and partners can use Moab’s graphical interface to select their preferred environment (hardware, OS, applications, etc.) and Moab will automatically provision the resources and give them secure access to the resources, increasing hardware sales and saving the time and costs associated with preparing test clusters.
- Allow customers with tight time frames access to your hardware the day the contract is signed and finalized.
Software Providers
- Host your high-end software to earn additional revenue from clients normally not able to purchase software licenses.
- Let customers easily access and test beta and production software to increase software sales.
Hosting Centers
- Create a seamless connection between yourself and your customers, allowing your hosted resources to immediately appear as an extension of your customer’s local cluster when they are needed.
- Automatically provision and prepare requested environments for your clients.
- Send out automated billing reports that charge based upon any resource consumption criteria such as CPU hours, license usage, bandwidth, storage, etc.
- Increase customer adoption by requiring no change in your customer’s job submission methods. Customers continue submitting standard scripts to their local cluster and Moab will automatically migrate jobs and data, translate needed scripts and provision the environment, requiring no change in end user experience.
Academic Sites
- Share expensive hardware and administrative costs by hosting your cluster to other departments or entities and send out automated billing based upon usage.
- Save money by renting expensive applications for short term projects.
Energy and Finance
- Ensure that resources are immediately available during workload surges without having to purchase resources that sit idle for long periods of time.
- Decrease staff distractions by letting them access hosted resources without changing job submission habits.
Government and Commercial Sites
- Keep all SLA and QoS guarantees by immediately accessing hosted resources when local resources fall short.
- Save money by renting resource, applications, or capacity and paying for only what you use instead of purchasing resources that sit idle.
| BASIC POLICIES |
||||
| Prioritization | ![]() |
![]() |
![]() |
![]() |
| Fairshare | ![]() |
![]() |
![]() |
![]() |
| Backfill | ![]() |
![]() |
![]() |
![]() |
| Reservations | ![]() |
![]() |
![]() |
![]() |
| Preemption | ![]() |
![]() |
![]() |
![]() |
| Job Management | ![]() |
![]() |
![]() |
![]() |
| Network Storage and License Policies | ![]() |
![]() |
||
| ACCOUNTING AND STATISTICS | ||||
| Basic Profiling (Credential, Node, Workload) |
![]() |
![]() |
![]() |
+ VPC |
| Advanced Profiling (Custom Metrics, Resources, Events) |
![]() |
![]() |
||
| Reports (Credential, Node, Workload) | ![]() |
![]() |
![]() |
+ VPC |
| Statistics (Credential, Node, Workload) | ![]() |
![]() |
![]() |
+ VPC |
| Grid Statistics and Reporting | ![]() |
![]() |
||
| Capacity Planning Reports | ![]() |
![]() |
||
| Per SLA/QOS Charging | ![]() |
![]() |
||
| GRID | ||||
| Moab to Moab Communication | ![]() |
![]() |
||
| Job Migration | ![]() |
|||
| Data Migration | ![]() |
![]() |
||
| Credential Mapping | ![]() |
![]() |
||
| Resource Manager Translation | ![]() |
![]() |
||
| WORKFLOW MANAGEMENT |
||||
| System Jobs | ![]() |
![]() |
||
| Workflow Support with Trigger Dependencies | ![]() |
![]() |
||
| Dynamic Virtual Service Jobs | ![]() |
![]() |
||
| Workload Performance Learning | ![]() |
![]() |
||
| Workload Health/Success Learning | ![]() |
![]() |
||
| Transaction ID Support | ![]() |
![]() |
||
| DYNAMIC RESOURCE MANAGEMENT |
||||
| Maximum Resource Manager Connections |
||||
| Integrated Virtualization | ![]() |
![]() |
||
| Node Provisioning | ![]() |
![]() |
||
| Holistic Resource Reporting/Accounting | ![]() |
![]() |
||
| Storage Management Integration | ![]() |
![]() |
||
| Network Management Integration | ![]() |
![]() |
||
| Full-environment Provisioning (Service, Security, Credential) |
![]() |
|||
| UTILITY COMPUTING |
||||
| SLA/QoS-based Performance Triggers | |
![]() |
||
| SOA Support | ![]() |
|||
| External RM Management Triggers | ![]() |
|||
| Provisioning Triggers | ![]() |
|||
| Virtual Private Clusters and Profiles | ![]() |
|||
| Unlimited Triggers, Partitions, Resource Manager Interfaces | ![]() |
|||
| Dynamic RM Management | ![]() |
|||
| Dynamic Workload and Environment Overflow to Other Systems | ![]() |
|||
| Dynamically Grow & Shrink Compute Resources for Workload or Services | ![]() |
|||
| INTERFACES |
||||
| End User Job Submission Portal | ![]() |
|
![]() |
![]() |
| Graphical Administrator Interface | ![]() |
![]() |
![]() |
![]() |
| Graphical Manager Interface | ![]() |
![]() |
![]() |
![]() |
| C API | ![]() |
![]() |
![]() |
![]() |
| XML | ![]() |
![]() |
![]() |
![]() |
| Java | ![]() |
![]() |
||
| Perl | ![]() |
|||


