Browse Prior Art Database

Method to manage reservation and allocation of scarce resources shared by tenants in a massively multi-tenant environment.

IP.com Disclosure Number: IPCOM000188508D
Original Publication Date: 2009-Oct-12
Included in the Prior Art Database: 2009-Oct-12
Document File: 6 page(s) / 141K

Publishing Venue

IBM

Abstract

Disclosed is a method to efficiently manage resource allocations to tenants in a massively multi-tenant environment, such as a Software-as-a-Service (SaaS) platform, in such a way that resource utilization is optimized; and the system is able to cater to changing market forces and business dynamics.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 22% of the total text.

Page 1 of 6

Method to manage reservation and allocation of scarce resources shared by tenants in a massively multi-tenant environment.

In a massively multi-tenant environment, such as a Software-as-a-Service (SaaS) environment, one of the major concerns is managing the shared resources. The SaaS Platform is required to manage the shared resources such as CPU, Memory, Hard-disk,

N

                                  ool among tenants efficiently. When the SaaS Provider on-boards a new tenant, the SaaS Platform reserve some resources based on the tenant-

        rofile, in order to satisfy the SLA with the tenant. Subsequently, the resources are incrementally provisioned on-demand.

In this context, disclosed is a method to efficiently manage resource allocations to tenants in a massively multi-tenant environment, such as a Software-as-a-Service (SaaS)

p

                                             latform, in such a way that resource utilization is optimized; and the system is able to cater to changing market forces and business dynamics. This problem can be further broken down into the following action:

Optimally reserve resources for a tenant (requesting for reserving resource) Optimally provision the reserved resources for a tenant (requesting for provisioning) Optimally re-allocate the reserved resources released by a tenant
Optimally re-allocate the provisioned resources released by a tenant

SaaS Platform reserve resources for a tenant based on the tenant's profile. SaaS Platform can dynamically change the amount of resource reserved for a tenant. SaaS Platform will provision resources to the tenant based on the resource reservation limit and /or availability. A tenant can request for resource of any size / amount. Multiple tenants can request for resources randomly at any point in time. A tenant tends to consume more and more resources as they consume a SaaS Service. A tenant may release resources randomly at any point in time. The market condition will affect the size of resource-

p

pool available for sharing. The SaaS Provider may dynamically adjust

        ool capacity levels. The market condition will also affect the rate at which the resource-

p

      ool will grow. The SaaS Provider may dynamically adjust the resource reservation levels

The characteristics of the 'tenant' consuming resources are as follows:

The tenants are classified /

profiled as platinum, gold, silver, bronze tenants depending on the

nature of business relationship with the SaaS Provider.

For example, Platinum > Gold > Silver > Bronze (

There is a well-defined 'resource reservation precedence' and 'resource allocation

p

rofile.

Precedence is defined using integer number ranging from [1, 10]. Higher the number, greater the precedence.

The characteristics of the 'resource ' managed by the SaaS Platform are as follows:

The resources are managed in the resource pool and the resource-

p

                                     ool is shared by multiple tenants. However a tenant cannot access the slice of resource provisioned to another

1

etwork, etc. in the

resource-

p

p...