SYSTEMS AND METHODS OF HOST-AWARE RESOURCE MANAGEMENT INVOLVING CLUSTER-BASED RESOURCE POOLS

• US 20140310710A1
• Filed: 06/27/2014
• Published: 10/16/2014
• Est. Priority Date: 02/22/2011
• Status: Active Grant

• Current Assignee: VIRTUSTREAM IP HOLDING COMPANY LLC (Dell Technologies Inc.)
• Original Assignee: Virtustream Inc. (Dell Technologies Inc.)
• Inventors: Reid, Kevin D., Rogers, Rodney John, LUBSEY, Vicent G., SIMPSON, Karl J.
• Granted Patent: US 9,535,752 B2
• Time in Patent Office: Days
• US Class Current: 718/1
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45583 Memory management, e.g. access or allocation
• CPC Class Codes: G06F 2009/45595 Network integration; Enabling network access in virtual machine instances
• CPC Class Codes: G06F 2209/5011 Pool
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: G06F 9/5072 Grid computing

Abstract

Systems and methods are disclosed for managing resources associated with cluster-based resource pool(s). According to illustrative implementations, innovations herein may include or involve one or more of best fit algorithms, infrastructure based service provision, tolerance and/or ghost processing features, dynamic management service having monitoring and/or decision process features, as well as virtual machine and resource distribution features.

Claims

1. A method for processing information related to management of a cluster-based pool of resources comprising one or more clusters wherein each cluster comprises one or more hosts executing one or more virtual machines (VMs), the method comprising:

establishing a customer'"'"'s resource requirement as a quantity of universal infrastructure units (IUs), wherein a universal infrastructure unit is defined as a function of processing speed and virtual machine memory size;

processing first information regarding configuration of the clusters with highwater marks to accomplish site and cluster tolerance, the highwater marks reflecting divisions between usable space and tolerance space, and between the tolerance space and host space reserved to ensure host activity;

processing second information regarding allocation of resource pools throughout the cluster usable space including handling of ghost pools located in unused tolerance space, the ghost pools being defined either on a host not containing a live pool associated with a customer or, when the customer has resource pools on all hosts in a cluster, on multiple hosts within the cluster;

processing third information regarding use of the ghost pools during maintenance windows or node tolerance events; and

processing fourth information regarding migration of virtual machines (VM'"'"'s) assigned to the customer'"'"'s resources back to originating pools, upon resumption of normal service.

2. The method of claim 1 further comprising:

analyzing clusters configured to hold customer data including performing speed testing and VM memory map sizing of hosts.

3. The method of claim 1 wherein, as a function of the allocation and associated failover processing, the VM'"'"'s of the customer are placed together in a limited subset of hosts, to provide faster communication of information and an established set of communication paths.

4. The method of claim 3 wherein, as a function of the allocation and associated failover processing, the cluster is configured such that maintenance windows need only to account for hosts associated with affected locations, rather than entirety of the cluster.

5. The method of claim 1 wherein, in connection with the allocation and associated failover processing, constraints are configured to remain in place even with loss of management service.

6. The method of claim 1 further comprising:

processing information regarding a quantity of infrastructure units for the customer, an indication of a number of virtual machines (VMs) for the customer, and an indication of a maximum number of VMs per cluster, the maximum number of VMs per cluster determined via calculation of the maximum number of VMs that achieve a guaranteed level of service corresponding to one or more predefined hardware resource limits.

7. The method of claim 6 wherein the one or more predefined hardware resource limits include a predefined outage time and the guaranteed level of service is IUs guaranteed to the customer, the maximum number of VMs per cluster being determined via calculation of the maximum number of VMs that achieve the predefined (or specified?) outage time established as a function of the IUs guaranteed to the customer.

8. The method of claim 7 wherein the predefined (or specified?) outage time is calculated as a function of the IUs guaranteed to the customer divided by migration or high availability time per virtual machine.

9. The method of claim 6 wherein guaranteed levels of service are provided for each predefined resource limit value from the set of predefined resource limit values.

10. The method of claim 1 wherein a value of the maximum number of virtual machines per cluster corresponds to a quantity of virtual machines set to ensure an appropriate speed for high availability (HA) restart or maintenance migration.

11. The method of claim 1 wherein the processing of information regarding migration of a VM includes providing room for a busy service within a cluster from which the VM is to be migrated.

12. The method of claim 1 wherein the processing of information regarding migration of a VM includes processing data regarding migration of the VM to a cluster having sufficient free space to handle operation of the VM task(s):

13. The method of claim 1 wherein the migration includes a customer-specific migration decision process that evaluates migration requirements, defines a best fit resolution plan, and performs processing to enact the best fit resolution plan, wherein the best fit resolution plan includes determination of a best fit of virtual machines across one or more hosts, to yield a specified outage time calculated as a function of infrastructure guarantee to the customer divided by either migration or high availability time per virtual machine.

14. The method of claim 13 further comprising performing processing to:

determine if any balance issue exists after the customer-specific migration decision process has performed its function(s); and

generate a new customer-specific migration decision process.

15. The method of claim 14 wherein the new customer-specific migration decision process results in a per-customer iterative balance.

18. The method of claim 6 wherein the one or more predefined hardware resource limits include a predefined outage time and the guaranteed level of service is IUs guaranteed to the customer, the maximum number of VMs per cluster being determined via calculation of the maximum number of VMs that achieve the predefined outage time established as a function of the IUs guaranteed to the customer.

19. The method of claim 18 wherein the predefined outage time is calculated as a function of the IUs guaranteed to the customer divided by migration or high availability time per virtual machine.

16. A method for processing information related to management of a cluster-based pool of resources comprising one or more clusters wherein each cluster comprises one or more hosts executing one or more virtual machines (VMs), the method comprising:

establishing a customer'"'"'s resource requirement as a quantity of universal infrastructure units (IUs), wherein a universal infrastructure unit is defined as a function of processing speed and virtual machine memory size;

processing information regarding a quantity of infrastructure units for the customer, an indication of a number of virtual machines (VMs) for the customer, and an indication of a maximum number of VMs per cluster, the maximum number of VMs per cluster determined via calculation of the maximum number of VMs that achieve a guaranteed level of service corresponding to one or more predefined hardware resource limits;

associating each infrastructure unit from the infrastructure units for the customer with hardware resources managed based on a set of predefined resource limit values assigned to the customer including operational parameters established for utilization of the customer'"'"'s resources while allocated among the one or more clusters;

calculating a number of clusters for the customer based on the number of infrastructure units for the customer, the number of VMs for the customer and the maximum number of VMs per cluster; and

processing information, based on the number of clusters for the customer, regarding distribution of VMs of the customer among a set of hosts of the one or more hosts such that each cluster for the customer is associated with VMs of the customer and is distributed to a unique host from the set of hosts.

17. The method of claim 16 further comprising:

analyzing clusters configured to hold customer data including performing speed testing and VM memory map sizing of hosts.

20. The method of claim 16 further comprising:

aggregating customer workload attributes; and

sending the aggregated information to a capacity planning engine or process.

21. The method of claim 16 further comprising:

performing allocation of a workload of the customer to the unique distribution of clusters in the cluster-based pool of resources.