Deploying virtual machine instances of a pattern to regions of a hierarchical tier using placement policies and constraints

• US 9,569,248 B2
• Filed: 04/11/2015
• Issued: 02/14/2017
• Est. Priority Date: 08/20/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Apte, Ajay A., Brabson, Roy F., Kirby, Theodore O., Moonen, Scott C., Woods, Donald R.
• Primary Examiner(s): Chen, Qing
• Application Number:
  US14/684,366
• Publication Number:
  US 20160055024A1
• Time in Patent Office: 675 Days
• Field of Search: 718/1, 717168-178
• US Class Current: 1/1
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45595 Network integration; Enabling network access in virtual machine instances
• CPC Class Codes: G06F 8/60 Software deployment
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: H04L 67/34 involving the movement of software or configuration parameters network booting or remote initial program loading [RIPL] G06F9/4416

Abstract

A method, system and computer program product for deploying a pattern. A placement engine of a hierarchical tier selects the region(s) of a next lower tier based on placement policies and constraints at the next lower tier and availability of the artifacts required by the virtual machine instance(s) in the next lower tier utilizing the summarization of resource usage and availability at the next lower tier. The placement engine at the lowest tier selects the compute node(s) in the selected region(s) of the lowest tier to place the virtual machine instances of the pattern based on placement policies and constraints at the compute node level and availability of the artifacts required by the virtual machine instance(s) of the pattern in the compute node level. In this manner, the virtual machine instances of the pattern can be effectively deployed when the resources of the cloud environment are large and geographically dispersed.

Claims

1. A method for deploying a pattern, the method comprising:

invoking, by a processor, a placement engine to place one or more virtual machine instances of said pattern in one or more compute nodes of a cloud computing environment;

selecting, by said processor, one or more regions in a second tier within one or more regions in a first tier to place said one or more virtual machine instances of said pattern based on placement policies and constraints at said second tier and availability of artifacts required by said one or more virtual machine instances of said pattern in said second tier;

sending, by said processor, a portion of said pattern to each of said selected one or more regions in said second tier to be deployed;

selecting, by said processor, one or more regions in a third tier within said selected one or more regions in said second tier to place said one or more virtual machine instances of said pattern based on placement policies and constraints at said third tier and availability of artifacts required by said one or more virtual machine instances of said pattern in said third tier;

sending, by said processor, a portion of said pattern to each of said selected one or more regions in said third tier to be deployed;

selecting, by said processor, one or more compute nodes in said selected one or more regions in said third tier to place said one or more virtual machine instances of said pattern based on placement policies and constraints at a compute node level and availability of artifacts required by said one or more virtual machine instances of said pattern in said compute node level; and

sending, by said processor, a portion of said pattern to each of said selected one or more compute nodes to be deployed.

2. The method as recited in claim 1 further comprising:

selecting said one or more regions in said second tier within said one or more regions in said first tier to place said one or more virtual machine instances of said pattern by a placement engine of a first tier of said placement engine;

selecting said one or more regions in said third tier within said selected one or more regions in said second tier by a placement engine of a second tier of said placement engine; and

selecting said one or more compute nodes in said selected one or more regions in said third tier to place said one or more virtual machine instances of said pattern by a placement engine of a third tier of said placement engine.

3. The method as recited in claim 1, wherein said one or more regions in said second tier comprises one or more sub-domains of a domain of said first tier, and wherein said one or more regions in said third tier comprises one or more racks of said one or more sub-domains of said second tier.

4. The method as recited in claim 1 further comprising:

aggregating resource usage and availability of resources of a current tier to form a summarization of tier resource usage and availability at said current tier; and

sending said summarization of said tier resource usage and availability at said current tier to a next higher tier.

5. The method as recited in claim 4 further comprising:

selecting said one or more regions in said second tier within said one or more regions in said first tier to place said one or more virtual machine instances of said pattern utilizing a summarization of resource usage and availability at said second tier;

selecting said one or more regions in said third tier within said selected one or more regions in said second tier utilizing a summarization of resource usage and availability at said third tier; and

selecting said one or more compute nodes in said selected one or more regions in said third tier utilizing a summarization of resource usage and availability at said compute node level.

6. The method as recited in claim 1, wherein said placement policies comprise placement restrictions for said one or more virtual machine instances of said pattern.

7. The method as recited in claim 1, wherein said constraints define relationships between various virtual machines in said pattern, and wherein said defined relationships determine whether said various virtual machines in said pattern are to be collocated or anti-collocated with respect to each other.