17643258. AUTOMATIC RESOURCE QUOTA CALCULATIONS BASED ON TENANT WORKLOADS simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

AUTOMATIC RESOURCE QUOTA CALCULATIONS BASED ON TENANT WORKLOADS

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Lior Aronovich of Thornhill (CA)

Kevin Doyle of Whitby (CA)

AUTOMATIC RESOURCE QUOTA CALCULATIONS BASED ON TENANT WORKLOADS - A simplified explanation of the abstract

This abstract first appeared for US patent application 17643258 titled 'AUTOMATIC RESOURCE QUOTA CALCULATIONS BASED ON TENANT WORKLOADS

Simplified Explanation

The abstract describes a method, computer program, and system for calculating resource quotas in a distributed computing environment. The system retrieves workload data for multiple workloads in a shared computing environment and identifies the tenants associated with these workloads. It determines the expected resource usage and ratio of resource usage for each tenant, and sets a resource limit for them. If the total expected resource usage for a tenant exceeds its resource limit, the system adjusts the shared computing environment accordingly.

  • The system calculates resource quotas in a distributed computing environment.
  • It retrieves workload data for multiple workloads in a shared computing environment.
  • It identifies the tenants associated with these workloads.
  • The system determines the expected resource usage and ratio of resource usage for each tenant.
  • It sets a resource limit for each tenant.
  • If a tenant's total expected resource usage exceeds its resource limit, the system makes adjustments to the shared computing environment.

Potential Applications

This technology can be applied in various scenarios where resource allocation and management are crucial in a distributed computing environment. Some potential applications include:

  • Cloud computing platforms: The system can be used to manage resource quotas for different tenants or users in a cloud computing environment, ensuring fair allocation and preventing resource abuse.
  • Data centers: It can be utilized to optimize resource allocation and prevent overutilization of resources in data centers, leading to improved efficiency and cost savings.
  • High-performance computing: The system can help allocate resources effectively among different users or projects in a high-performance computing environment, ensuring optimal performance and resource utilization.

Problems Solved

This technology addresses several problems in resource management and allocation in a distributed computing environment, including:

  • Fairness: By setting resource quotas based on expected usage and adjusting the environment when limits are exceeded, the system ensures fair allocation of resources among different tenants or users.
  • Resource abuse prevention: By monitoring and limiting resource usage, the system prevents any single tenant from monopolizing resources and ensures that resources are available for others.
  • Optimization: The system optimizes resource allocation by considering expected usage and adjusting the environment accordingly, leading to improved efficiency and performance.

Benefits

The use of this technology offers several benefits in a distributed computing environment:

  • Fair and equitable resource allocation: By setting resource quotas and adjusting the environment when necessary, the system ensures that all tenants or users have access to resources in a fair and equitable manner.
  • Resource optimization: The system optimizes resource allocation based on expected usage, preventing overutilization and ensuring efficient resource utilization.
  • Preventing resource abuse: By monitoring and limiting resource usage, the system prevents any single tenant from monopolizing resources, promoting a balanced and sustainable computing environment.


Original Abstract Submitted

A method, computer program product and computer system to calculate resource quotas in a distributed computing environment are provided. A processor retrieves workload data regarding a plurality of workloads in a shared computing environment, where the plurality of workloads are executing or pending execution within the shared computing environment. A processor identifies a plurality of tenants of the shared computing environment associated with the plurality of workloads. A processor determines an expected resource usage for the plurality of tenants. A processor determines a ratio of resource usage for the plurality of tenants. A processor determines a resource limit for the plurality of tenants. A processor adjusts at least one aspect of the shared computing environment based on a determination that a total expected resource usage for both executing and pending workloads of a tenant exceeds a resource limit associated with the tenant.