Method and system for performing load balancing across control planes in a data center

• US 20040117476A1
• Filed: 12/17/2002
• Published: 06/17/2004
• Est. Priority Date: 12/17/2002
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Hogan, Katherine, Schloss, Rheid, Steele, Doug
• Granted Patent: US 7,933,983 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

A system and method analyzes resource usage within a utility data center (UDC) and recommends options to maximize efficient usage of these resources. Information is gathered about the current resources allocated to control planes within the UDC, including resource loads and threshold limits within each plane. Criteria used to analyze efficiency includes customer size and prioritization within the UDC, as well as, historical and trending information, and peak resource usage.

Claims

1. A system for load balancing network components in a data center, comprising:

at least one enterprise system executing on a virtual local area network (VLAN), the VLAN comprising a plurality of network components;

at least one control plane management system in a control plane that controls one or more VLANs utilizing a protocol for uniquely identifying each network component of the one or more VLANs controlled by the control plane management system, wherein network communication within the control plane is routed to a respective VLAN component based on a respective unique identifier, and wherein the control plane management system initially allocates respective network components to a specified VLAN based on requirements of each enterprise system;

means for collecting performance metrics from selected network components of the plurality of network components; and

means for performing load analysis using the collected performance metrics and pre-determined performance thresholds, the analysis means resulting in a recommended allocation and configuration of network components in the control plane, wherein the control plane automatically reconfigures network components based on the recommended allocation.

2. The system as recited in claim 1, wherein the control plane automatically reconfigures network components based on the recommended allocation, only if such reallocation is authorized.

3. The system as recited in claim 1, wherein the means for collecting performance metrics uses, for each of the selected network components, at least one standard support-level monitoring tool corresponding to the respective selected component.

4. The system as recited in claim 3, wherein monitoring tools include at least one tool selected from the group consisting of:

CPU usage monitor, disk usage monitor, throughput monitor, Input/output (I/O) monitor, and network traffic monitor.

5. The system as recited in claim 1, further comprising:

a network operations center (NOC) which oversees the at least one control plane management system in the data center, the NOC communicating with each control plane management system through one or more firewalls and communicating with the VLAN components through its corresponding control plane management system and via at least one predefined port corresponding to the protocol for uniquely identifying each component, wherein the NOC utilizes a standard port and deploys local agents using translators for existing support actions, wherein support services are deployed to the at least one control plane management system and each of the at least one control plane management system and each of the at least one enterprise to handle request and event objects sent through the firewalls;

means for collecting network performance metrics from selected network components of the data center; and

means for load analysis across control planes using the collected network performance metrics and pre-determined performance thresholds, the network analysis means resulting in a recommended network allocation and configuration of components in the data center.

6. The system as recited in claim 4, wherein an operator reconfigures resources across control planes in the data center, thereby reconfiguring network components based on the recommended allocation.

7. The system as recited in claim 6, wherein an operator reconfigures resources across control planes in the data center, only if such reallocation is authorized.

8. The system as recited in claim 6, wherein an enterprise owner is notified of the recommended network allocation before reconfiguring.

9. The system as recited in claim 1. wherein pre-determined performance thresholds are selected in a service level agreement between the enterprise owner and an owner/operator of the data center.

10. A method for automatic load balancing of resources among enterprises in a data center, comprising:

deploying at least one enterprise system executing on a virtual local area network (VLAN), wherein the VLAN comprises a plurality of network components, and wherein at least one control plane management system in a control plane controls one or more VLANs, the control plane utilizing a protocol for uniquely identifying each network component of the one or more VLANs controlled by the control plane management system, and wherein network communication within the control plane is routed to a respective VLAN component based on the unique identifier, and wherein the control plane management system allocates each network component to a specified VLAN based on requirements of each enterprise system, and wherein a network operation center (NOC) oversees the at least one control plane management system in the data center;

collecting performance metrics from selected network components of the plurality of network components; and

performing load analysis using the collected performance metrics and pre-determined performance thresholds, the load analysis resulting in a recommended allocation and configuration of network components in the control plane; and

automatically reconfiguring network components, by the control plane manager, based on the recommended allocation.

11. The method as recited in claim 10, wherein automatically reconfiguring network components, by the control plane manager is performed, only if such reallocation is authorized.

12. The method as recited in claim 10, further comprising:

communicating, by the NOC, with each control plane management system and to the VLAN components through its corresponding control plane management system, the communicating occurring through one or more firewalls via at least one predefined port corresponding to the protocol for uniquely identifying each component, wherein the NOC utilizes a standard port for communication through the firewall to deploy local agents using translators for existing support actions; and

deploying support-level monitoring tools to the at least one control plane management system and the at least one enterprise through the firewalls, wherein support-level monitoring tools are deployed to each of the at least one control plane management system and each of the at least one enterprise to handle collection requests and event objects sent through the firewalls.

13. The method as recited in claim 12, further comprising:

collecting network performance metrics from selected network components of the data center; and

performing network load analysis across control planes using the collected network performance metrics and pre-determined performance thresholds, wherein the network load analysis results in a recommended network allocation and configuration of components in the data center.

14. A system for automatically load balancing network components in a data center, comprising:

at least one enterprise system executing on a virtual local area network (VLAN), the VLAN comprising a plurality of network components;

at least one control plane management system in a control plane that controls one or more VLANs utilizing a protocol for uniquely identifying each network component of the one or more VLANs controlled by the control plane management system, wherein the control plane management system initially allocates each network component in the control plane to a specified VLAN based on requirements of each enterprise system;

means for collecting performance metrics from selected network components of the plurality of network components, wherein at least one monitoring tool for a selected component is installed on the network component when the network component is deployed in the data center; and

means for performing load analysis using the collected metrics and pre-determined performance thresholds, the analysis means resulting in a recommended allocation and configuration of components in the control plane.

15. The system as recited in claim 14, wherein the control plane automatically reconfigures network components controlled by the control plane, based on the recommended allocation, if such reallocation is authorized, and wherein reallocation is not authorized, automatically sending a notification with the recommended allocation.

16. The system as recited in claim 14, wherein a recommendation for a first control plane reallocation is compared with a recommendation for a second control plane reallocation to determine whether resources should be reallocated between the first and second control plane.

17. The system as recited in claim 14, wherein the recommended allocation of resources recommends deployment of additional network components in a control plane to alleviate loading of current control plane resources.

18. The system as recited in claim 14, wherein means for collecting performance metrics use, for each selected network component, at least one standard support-level monitoring tool corresponding to the each selected network component.

19. The system as recited in claim 18, wherein monitoring tools include at least one tool selected from the group consisting of:

CPU usage monitor, disk usage monitor, throughput monitor, Input/output (I/O) monitor, and network traffic monitor.

20. The system as recited in claim 14, further comprising:

a network operations center (NOC) which oversees the at least one control plane management system in the data center, the NOC communicating with each control plane management system through one or more firewalls and communicating with the VLAN components through its corresponding control plane management system and via at least one predefined port corresponding to the protocol for uniquely identifying each component;

means for collecting network performance metrics from selected network components of the data center; and

means for load analysis across control planes using the collected network metrics and pre-determined performance thresholds, the network analysis means resulting in a recommended network allocation and configuration of network components in the data center.

21. A load balancer for a data center, the data center having a network operation center, one or more control planes, and one or more enterprises within a control plane, comprising:

means for collecting performance measurements at an enterprise resource level, wherein measurements are available at various levels of a data center hierarchy, the hierarchy comprising;

a network operation center (NOC) having at least one control plane, and at least one enterprise running in a virtual local area network (VLAN) within each control plane, wherein the control planes are monitored and further controlled by the NOC through various firewalls;

means for comparing respective performance measurements of a plurality of respective enterprises, wherein the comparison means analyzes the performance measurements and recommends further action; and

means for performing the recommended further action by one of a control plane management system and a NOC operator.

22. The load balancer of claim 21, wherein the recommended further action is performed automatically by the control plane management system.

23. The load balancer of claim 21, wherein the recommended further action is a reallocation of control plane resources.

24. The load balancer of claim 21, wherein the recommended further action is a reallocation of data center resources across control planes.

25. The load balancer of claim 21, wherein the recommended further action is a reallocation of data center network resources.