Dynamically building a set of compute nodes to host the user's workload

Patent metadata

Field	Value
Publication number	US-8930543-B2
Application number	US-201313858849-A
Country	US
Kind code	B2
Filing date	Apr 8, 2013
Priority date	Jan 23, 2012
Publication date	Jan 6, 2015
Grant date	Jan 6, 2015

Abstract

A method, system and computer program product for dynamically building a set of compute nodes to host a user's workload. An administrative server receives workload definitions that include the types of workloads that are to be run in a cloud group as well as a number of instances of each workload the cloud group should support. These workload definitions are used to determine the virtual machine demands that the cloud group will place on the cloud environment. The administrative server further receives the demand constraints, placement constraints and license enforcement policies. The administrative server identifies a set of compute nodes to host the user's workload based on the virtual machines demands, the demand constraints, the placement constraints and the license enforcement policies. In this manner, a set of compute nodes is dynamically built for consideration in forming a cloud group without the user requiring knowledge of the cloud's composition.

First claim

The invention claimed is:

1. A method for dynamically building a set of compute nodes to host a user's workload, the method comprising:
receiving workload definitions comprising types of workloads that are to be run in a cloud group as well as a number of instances of each workload said cloud group should support;
using said workload definitions to determine virtual machine demands that said cloud group will place on a cloud computing environment;
receiving demand constraints on said cloud group;
receiving placement constraints on said cloud group; and
identifying, by a processor, said set of compute nodes to host said user's workload based on said virtual machine demands, said demand constraints and said placement constraints.

2. The method as recited in claim 1 , wherein said demand constraints on said cloud group comprise one or more of the following: processor, memory, storage, network I/O, storage I/O and bandwidth constraints.

3. The method as recited in claim 1 , wherein said placement constraints on said cloud group comprise one or more of the following: high availability, consolidation and energy conservation constraints.

4. The method as recited in claim 1 further comprising:
receiving license enforcement policies; and
identifying said set of compute nodes to host said user's workload based on said virtual machine demands, said demand constraints, said placement constraints and said license enforcement policies.

5. The method as recited in claim 1 further comprising:
deploying virtual machines to said set of compute nodes; and
deploying workloads to said set of compute nodes.

6. The method as recited in claim 5 further comprising one or more of the following:
monitoring for changes in workload demands, said demand constraints and said placement constraints after said deployment of said virtual machines;
monitoring for hardware failures and for predicted hardware failures after said deployment of said virtual machines;
monitoring demands on said virtual machines after said deployment of said virtual machines;
monitoring for hardware usage after said deployment of said virtual machines;
monitoring for additions or subtractions of hardware in said cloud group after said deployment of said virtual machines; and
monitoring for license usage after said deployment of said virtual machines.

7. The method as recited in claim 6 further comprising one or more of the following:
rebalancing computing resources in said cloud group in response to one or more of said workload demands, said demand constraints and said placement constraints changing after said deployment of said virtual machines;
said rebalancing of said computing resources in said cloud group in response to detecting said hardware failure or said predicted hardware failure;
said rebalancing of said computing resources in said cloud group in response to one or more virtual machines of said deployed virtual machines not receiving required demands;
said rebalancing of said computing resources in said cloud group in response to hardware not being fully utilized;
said rebalancing of said computing resources in said cloud group in response to said additions or subtractions of hardware in said cloud group; and
said rebalancing of said computing resources in said cloud group in response to a license being available.

8. The method as recited in claim 7 , wherein said rebalancing of said computing resources comprises one or more of the following:
relocating one or more virtual machines in said deployed virtual machines, wherein said relocating comprises chained relocations;
adding one or more virtual machines to said deployed virtual machines in said set of compute nodes; and
relocating one or more virtual machines in said deployed virtual machines in said set of compute nodes of said cloud group to other compute nodes outside of said cloud group.

Assignees

• IBM

Inventors

• Ashok Rohith K
• Brabson Roy F
• Hockett Hugh E
• Hogstrom Matt R
• Segmuller Wolfgang
• Sheard Matthew J

Classifications

• G06F9/5072 Primary

  Grid computing · CPC title

• H04L67/1031 Primary

  Controlling of the operation of servers by a load balancer, e.g. adding or removing servers that serve requests · CPC title