Method for managing resources in a platform for telecommunication service and/or network management, corresponding platform and computer program product therefor

• US 8,264,971 B2
• Filed: 10/28/2004
• Issued: 09/11/2012
• Est. Priority Date: 10/28/2004
• Status: Active Grant

• Current Assignee: Telecom Italia S.p.A.
• Original Assignee: Telecom Italia S.p.A.
• Inventors: Alfano, Rosario, Cassone, Giuseppe, Covino, Giuseppe, Gotta, Danilo, Porta, Marisa, Bobbio, Fabrizio, Ughetti, Marco
• Primary Examiner(s): Vu, Huy D
• Assistant Examiner(s): MIAN, OMER S
• Application Number:
  US11/666,509
• Publication Number:
  US 20090122706A1
• Time in Patent Office: 2,875 Days
• Field of Search: 370/252, 370/250, 370/242, 709/224, 709/226
• US Class Current: 370/242
• CPC Class Codes: H04L 41/046 comprising network management agents or mobile agents therefor
• CPC Class Codes: H04L 41/5003 Managing SLA; Interaction between SLA and QoS
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04Q 3/0087 Network testing or monitoring arrangements

Abstract

A method and system for managing resources in telecommunications services and networks. The method and system regard execution of processes by distributed agents on processing machines to perform predetermined tasks on the network, and are characterized by setting target data including goals on processes and constraints on resource utilization to be met by the system; monitoring computational resources utilization and processes execution by agents, and collecting performance data representative of resource utilization and process execution; comparing the collected performance data with the set data, a penalty being established when performance data do not fulfill target data; and re-allocating resources for process execution by agents so as to minimize the penalty determined on the basis of the comparison.

Claims

1. A method for managing resources in a platform for at least one of a telecommunication service and network management, wherein the platform is able to manage process executions having process priorities executed by distributed agents, comprising:

setting target data to be met by the platform, wherein said target data comprises goals on the process executions executed by the distributed agents and constraints on platform resource utilization;

monitoring, using a local performance control agent associated with a plurality of distributed agents and co-located with the associated plurality of distributed agents on a host computer, the process executions executed by each distributed agent and resource utilization associated with the process executions performed by each distributed agent, wherein each distributed agent independently executes process executions spontaneously;

performing, by the local performance control agent, local optimization of resource management;

collecting from the local performance control agent performance data representative of said process executions and of said resource utilization;

comparing the collected performance data with the set target data;

establishing at least one penalty clause on the basis of a comparison between the collected performance data of the distributed agents and the set target data; and

re-allocating resources to distributed agents to execute process executions based in part on the established at least one penalty clause, wherein re-allocating resources comprises;

executing an evaluation step and a decision step at determined observation intervals, wherein as part of the evaluation step, the collecting step includes,

collecting from the local performance control agent, data representative both of process executions and of the number of forecasted process executions for at least one of subsequent observation intervals, and evaluating, on the basis of said collected data, the resources as required by said distributed agents, and as part of the decision step, the comparing step includes

comparing the resources as required to resources as available by each of said distributed agents, and applying determined resource re-allocation rules to said distributed agents in order to modify resource utilization among at least one of the distributed agents, to change process priorities at the distributed agents, and to reallocate process executions among the distributed agents.

2. The method according to claim 1, comprising:

providing distributed process engines to the distributed agents; and

storing process descriptions representative of processes in a process descriptions database associated with said distributed process engines.

3. The method according to claim 2, wherein the process descriptions comprise at least one of workflows and rules.

4. The method according to claim 1, comprising:

providing the distributed agents on hierarchical levels according to a multi-layered configuration of agents.

5. The method according to claim 4, wherein the process executions are assigned to the multi-layered configuration of distributed agents by a centralized manager module.

6. The method according to claim 5, wherein the collecting performance data step comprises:

transmitting said performance data to at least one of the centralized manager module and to a plurality of local performance control agents associated with the distributed agents.

7. The method according to claim 5, comprising:

providing at least one master agent at the top layer of said multi-layered configuration of distributed agents, the master agent charging process executions to distributed agents located in sub-layers of said multi-layered configuration.

8. The method according to claim 1, comprising, for each distributed agent, the steps of:

inserting a process execution request into a multilevel priority process queue according to a priority criterion; and

scheduling the process executions on the basis of the multilevel priority process queue.

9. The method according to claim 8, comprising scheduling process executions by at least one process engine thread associated with each distributed agent.

10. The method according to claim 8, wherein the process execution requests in the multilevel priority process queue are upgraded based on a timeout criterion.

11. The method according to claim 6, wherein each control agent controls a number of process engine threads and resource utilization by the distributed agents.

12. The method according to claim 6, wherein

the control agent executes a load balancing algorithm for determining the load of the distributed agents;

and

each distributed agent sends process execution requests to other distributed agents on the basis of a criterion, including at least an evaluation of the load of the distributed agents as determined by the control agent.

26. A method for setting up and managing telecommunication services, or an ADSL (Asymmetric Digital Subscriber Line) connectivity service, comprising the step of managing resources according to the method as defined in claim 1.

27. A non-transitory computer readable medium encoded with a computer program product or computer program set of computer program products loadable into a memory of at least one computer, the computer program product or computer program set of computer program products comprising software code portions for performing the steps of claim 1.

13. A platform for managing resources for at least one of telecommunication services and networks, comprising

a plurality of distributed agents able to manage process executions having process priorities, comprising:

process engines associated with said distributed agents and configured for independently executing the process executions spontaneously and monitoring the process executions and resource utilization by the distributed agents, and by

a centralized manager module, configured for;

setting target data to be met by the platform, wherein said target data comprises goals on the process executions by the distributed agents and constraints on platform resource utilization to be met by the platform;

collecting performance data representative of said process executions and of said resource utilization by the distributed agents;

comparing the collected performance data with the set target data;

establishing at least one penalty clause on the basis of a comparison between the collected performance data of said distributed agents and the set target data;

re-allocating resources to said distributed agents for process executions by distributed agents on the basis of said established at least one penalty clause; and

a local performance control agent associated with at least two distributed agents of the plurality of distributed agents, wherein the local performance control agent is co-located with the at least two associated distributed agents on a host computer, and the local performance control agent is configured to perform local optimization of resource management,

wherein said centralized manager module comprises a resource allocator module comprising;

an evaluate module configured for;

collecting, from the local performance control agent, data representative both of process executions and of the number of forecasted process executions for a subsequent observation interval, and evaluating, on the basis of said collected data, the resources as required by said distributed agents, and

a decide module configured for;

comparing the resources as required to resources as available by each of said distributed agents, and applying determined resource re-allocation rules to said distributed agents in order to modify at least one of resource utilization among the distributed agents, to change process priorities at the distributed agents, and to reallocate process executions among the distributed agents.

14. The platform according to claim 13, wherein said centralized manager module comprises:

a process description database for storing process descriptions representative of behavorial and functional aspects of the platform.

15. The platform according to claim 14, wherein said centralized manager module further comprises:

a service creation console arranged for definition, creation and modification of the process descriptions in the process description database.

16. The platform according to claim 14, wherein the process descriptions comprise workflows and/or rules.

17. The platform according to claim 13, wherein

said plurality of distributed agents are organized on hierarchical levels according to a multi-layered configuration, and wherein

said centralized manager module is configured for assigning process executions to said multi-layered configuration of distributed agents.

18. The platform according to claim 13, comprising:

local performance control agents associated with at least a set of distributed agents, said process engines comprising resource monitor modules being configured for;

transmitting said performance data to the centralized manager module and/or to the local performance control agents associated with the distributed agents.

19. The platform according to claim 17, comprising:

at least one master agent located at the top layer of said multi-layered configuration of distributed agents and configured for charging process executions to distributed agents located in sub-layers of said multi-layered configuration.

20. The platform according to claim 13, comprising:

at least one processing machine comprising at least one set of said plurality of distributed agents.

21. The platform according to claim 20, wherein at least one local performance control agent is associated with said at least one processing machine.

22. The platform according to claim 21, wherein said at least one local performance control agent comprises:

a common local performance monitoring module arranged for collecting performance data representative of resource utilization and process execution by the distributed agents and transmitting the performance data to the centralized manager module;

a common thread controller coupled to the resource monitor arranged for creating process engine threads for executing waiting processes; and

a common dispatcher module coupled to the process engine threads and arranged for sending process execution requests to other distributed agents according to a predetermined load balancing algorithm.

23. The platform according to claim 13, wherein the manager module comprises:

a capacity planning module configured for;

forecasting availability of resources in an observation interval on the basis of historical performance and of data representative of present resource utilization.

24. The platform according to claim 13, wherein the manager module comprises:

an administrative console configured for;

defining the hardware configuration of the platform; and

defining constraints on process executions.

25. A telecommunications network managed by a platform according to claim 13.

28. A method for managing resources in at least one of a telecommunication service and network, the method comprising:

running a plurality of distributed agents, the distributed agents including at least one process engine for executing at least one management process, wherein the at least one process engine includes at least one of a workflow engine and a rule engine;

setting target data to be met, the target data including goals on the process executions executed by the plurality of distributed agents and constraints on resource utilization;

monitoring, by the at least one process engine, the process executions executed by each distributed agent and resource utilization associated with the process executions performed by each distributed agent, wherein each distributed agent independently executes process executions spontaneously;

performing, by a local performance control agent associated with, at least two distributed agents of the plurality of distributed agents, local optimization of resource management, wherein the local performance control agent and the at least two distributed agents are co-located on a host computer;

collecting performance data representative of said process executions and of said resource utilization;

comparing the collected performance data with the set target data;

establishing at least one penalty clause on the basis of a comparison between the collected performance data of the distributed agents and the set target data; and

re-allocating resources to distributed agents to execute process executions based in part on the established at least one penalty clause, wherein re-allocating resources comprises;

executing an evaluation step and a decision step at determined observation intervals, wherein as part of the evaluation step,

the collecting step includes,

collecting from the local performance control agent, data representative both of process executions and of the number of forecasted process executions for at least one of subsequent observation intervals, and evaluating, on the basis of said collected data, the resources as required by said distributed agents, and

as part of the decision step, the comparing step includes

comparing the resources as required to resources as available by each of said distributed agents, and applying determined resource re-allocation rules to said distributed agents in order to modify resource utilization among at least one of the distributed agents, to change process priorities at the distributed agents, and to reallocate process executions among the distributed agents.

29. A platform for managing resources for at least one of telecommunication services and networks, comprising

a plurality of distributed agents, each distributed agent including at least one process engine capable of managing the execution of at least one management process, the at least one process engine including at least one of a workflow engine and a rule engine, and monitoring the process executions and resource utilization by the distributed agents, wherein each distributed agent independently executes process executions spontaneously;

and

a centralized manager module, configured for;

setting target data to be met by the platform, wherein said target data includes goals on the process executions by the distributed agents and constraints on platform resource utilization to be met by the platform;

collecting performance data representative of said process executions and of said resource utilization by the distributed agents;

comparing the collected performance data with the set target data;

establishing at least one penalty clause on the basis of a comparison between the collected performance data of said distributed agents and the set target data;

re-allocating resources to said distributed agents for process executions by said distributed agents on the basis of said established at least one penalty clause; and

a local performance control agent associated with at least two distributed agents of the plurality of distributed agents, wherein the local performance control agent is co-located with the at least two associated distributed agents on a host computer, and the local performance control agent is configured to perform local optimization of resource management,

wherein said centralized manager module comprises a resource allocator module comprising;

an evaluate module configured for;

collecting, from the local performance control agent, data representative both of process executions and of the number of forecasted process executions for a subsequent observation interval, and

evaluating, on the basis of said collected data, the resources as required by said distributed agents, and

a decide module configured for;

comparing the resources as required to resources as available by each of said distributed agents, and

applying determined resource re-allocation rules to said distributed agents in order to modify at least one of resource utilization among the distributed agents, to change process priorities at the distributed agents, and to reallocate process executions among the distributed agents.