Scheduler for use in a scalable, distributed, asynchronous data collection mechanism

• US 6,421,676 B1
• Filed: 06/30/1999
• Issued: 07/16/2002
• Est. Priority Date: 06/30/1999
• Status: Expired due to Term

• Current Assignee: HGST Netherlands B.V. (Western Digital Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Mallo, Michael, Krishnamurthy, Raghavendra
• Primary Examiner(s): Black, Thomas
• Assistant Examiner(s): Mofiz, Apu M
• Application Number:
  US09/345,628
• Time in Patent Office: 1,112 Days
• Field of Search: 709/102, 709/103, 709/106, 709/201, 709/205, 709/206, 709/217-220, 709/223, 709/224, 709/225, 709/229, 709/310, 709/313, 709/319, 707/102, 712/28, 712/25, 712/201, 712/245
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2471 Distributed queries
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Scheduling in a distributed data collection process is performed locally, within collectors. Scheduling of data transfers from endpoints or downstream collectors or to upstream collectors is based on local queues, without global management. Additionally, scheduling for the collector input queue, which manages data collection from endpoints or downstream collectors, is bifurcated from scheduling for the output queue, which manages notifications to upstream collector(s) regarding the availability of collection data for pickup. Such bifurcation permits simpler scheduling logic and different functional responses to similar events, and further localizes scheduling. Scheduling of collection data transfer is controlled, within parameters specified by the output scheduler for the endpoint or downstream collector, by the input queue for the upstream collector. Scheduling is thus based primarily on the portion of the data transfer mechanism mostly likely to comprise a bottle-neck, the upstream collector, but accommodates large numbers of fully parallel data generation endpoints as well as nondeterministic endpoint availability.

Claims

1. A method of scheduling distributed data collection, comprising:

scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector, wherein the input queue contains data structures corresponding to all collection data available at an endpoint or downstream collector for transfer by the collector; and

responsive to transfer of collection data from an endpoint or downstream collector to the collector, notifying an upstream collector of availability of the collection data at the collector.

2. The method of claim 1, wherein the step of notifying an upstream collector of availability of the collection data at the collector further comprises:

setting an activation time window within a data structure corresponding to the collection data.

3. The method of claim 2, further comprising:

utilizing the activation time window to define black out periods for a node containing the collector.

4. The method of claim 2, further comprising:

utilizing the activation time window for bandwidth utilization control and to handle endpoints intermittently available for data collection.

5. The method of claim 1, further comprising:

providing a customization and control module within the collector to support status monitoring and the ability to stop the collector and drain queues within the collector.

6. The method of claim 1, further comprising:

providing a graphical topological view of a collection network including the collector for monitoring the collection network.

7. The method of claim 1, wherein the step of scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector further comprises:

scheduling the data collection based on local queues, independent global timing control.

8. The method of claim 1, wherein the step of scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector further comprises:

scheduling the data collection based on separate input and output queues, wherein the input queue is utilized for data collection from a downstream endpoint or collector and the output queue is utilized for data transfer to an upstream collector or recipient.

9. A method of scheduling distributed data collection, comprising:

receiving a data structure indicating that collection data is available at an endpoint or downstream collector;

placing the data structure within an input queue for a collector receiving the data structure;

sorting the data structure and any other data structures within the input queue based on preselected attributes of the data structures;

responsive to resources at an upstream collector for transferring the collection data becoming available, selecting a next data structure from the sorted data structures within the input queue;

responsive to successful transfer of the collection data, removing the data structure from the input queue and placing the data structure in an output queue for the collector.

10. The method of claim 9, further comprising:

providing the collector with a plurality of modes of operation,

wherein a first mode of operation, corresponding to a first priority level within a collection table of contents data structure for collection data, is employed for collecting data from mobile endpoints, and

wherein a second mode of operation, corresponding to a second priority level within the collection table of contents data structure for collection data, is employed for collecting data from continuously available endpoints.

11. A system for scheduling distributed data collection, comprising:

an input queue containing data structures corresponding to all collection data available at an endpoint or downstream collector for transfer by a collector;

a scheduler scheduling data collection by the collector from an endpoint or downstream collector based on contents of the input queue for the collector; and

signal means, responsive to transfer of collection data from an endpoint or downstream collector to the collector, for notifying an upstream collector of availability of the collection data at the collector.

12. The system of claim 11, wherein the signal means for notifying an upstream collector of availability of the collection data at the collector further comprises:

means for setting an activation time window within a data structure corresponding to the collection data.

13. The system of claim 12, further comprising:

means for utilizing the activation time window to define black out periods for a node containing the collector.

14. The system of claim 12, further comprising:

means for utilizing the activation time window for bandwidth utilization control and to handle endpoints intermittently available for data collection.

15. The system of claim 11, further comprising:

means for providing a customization and control module within the collector to support status monitoring and the ability to stop the collector and drain queues within the collector.

16. The system of claim 11, further comprising:

means for providing a graphical topological view of a collection network including the collector for monitoring the collection network.

17. The system of claim 11, wherein the scheduler further comprises:

means for scheduling the data collection based on local queues, independent global timing control.

18. The system of claim 11, wherein the sscheduler further comprises:

means for scheduling the data collection based on separate input and output queues, wherein the input queue is utilized for data collection from a downstream endpoint or collector and the output queue is utilized for data transfer to an upstream collector or recipient.

22. The computer program product of claim 11, wherein the instructions for notifying an upstream collector of availability of the collection data at the collector further comprises:

instructions for setting an activation time window within a data structure corresponding to the collection data.

23. The computer program product of claim 22, further comprising:

instructions for utilizing the activation time window to define black out periods for a node containing the collector.

24. The computer program product of claim 22, further comprising:

instructions for utilizing the activation time window for bandwidth utilization control and to handle endpoints intermittently available for data collection.

19. A system for scheduling distributed data collection, comprising:

means for receiving a data structure indicating that collection data is available at an endpoint or downstream collector;

means for placing the data structure within an input queue for a collector receiving the data structure;

means for sorting the data structure and any other data structures within the input queue based on preselected attributes of the data structures;

means, responsive to resources at an upstream collector for transferring the collection data becoming available, for selecting a next data structure from the sorted data structures within the input queue;

means, responsive to successful transfer of the collection data, for removing the data structure from the input queue and placing the data structure in an output queue for the collector.

20. The system of claim 19, further comprising:

means for providing the collector with a plurality of modes of operation,

wherein a first mode of operation, corresponding to a first priority level within a collection table of contents data structure for collection data, is employed for collecting data from mobile endpoints, and

wherein a second mode of operation, corresponding to a second priority level within the collection table of contents data structure for collection data, is employed for collecting data from continuously available endpoints.

21. A computer program product within a computer usable medium for scheduling distributed data collection, comprising:

instructions for scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector, wherein the input queue contains data structures corresponding to all collection data available at an endpoint or downstream collector for transfer by the collector; and

instructions, responsive to transfer of collection data from an endpoint or downstream collector to the collector, for notifying an upstream collector of availability of the collection data at the collector.

25. The computer program product of claim 21, further comprising:

instructions for providing a customization and control module within the collector to support status monitoring and the ability to stop the collector and drain queues within the collector.

26. The computer program product of claim 21, further comprising:

instructions for providing a graphical topological view of a collection network including the collector for monitoring the collection network.

27. The computer program product of claim 21, wherein the instructions for scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector further comprise:

instructions for scheduling the data collection based on local queues, independent global timing control.

28. The computer program product of claim 21, wherein the instructions for scheduling data collection for a collector from an endpoint or downstream collector based on contents of an input queue for the collector further comprise:

instructions for scheduling the data collection based on separate input and output queues, wherein the input queue is utilized for data collection from a downstream endpoint or collector and the output queue is utilized for data transfer to an upstream collector or recipient.

29. A computer program product within a computer usable medium for scheduling distributed data collection, comprising:

instructions for receiving a data structure indicating that collection data is available at an endpoint or downstream collector;

instructions for placing the data structure within an input queue for a collector receiving the data structure;

instructions for sorting the data structure and any other data structures within the input queue based on preselected attributes of the data structures;

instructions, responsive to resources at an upstream collector for transferring the collection data becoming available, for selecting a next data structure from the sorted data structures within the input queue;

instructions, responsive to successful transfer of the collection data, for removing the data structure from the input queue and placing the data structure in an output queue for the collector.

30. The computer program product of claim 29, further comprising:

instructions for providing the collector with a plurality of modes of operation,

wherein a first mode of operation, corresponding to a first priority level within a collection table of contents data structure for collection data, is employed for collecting data from mobile endpoints, and

wherein a second mode of operation, corresponding to a second priority level within the collection table of contents data structure for collection data, is employed for collecting data from continuously available endpoints.