Scaling event processing using distributed flows and map-reduce operations

• US 10,120,907 B2
• Filed: 09/22/2015
• Issued: 11/06/2018
• Est. Priority Date: 09/24/2014
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: de Castro Alves, Alexandre, Thukkaram, Prabhu, Bishnoi, Sandeep, Bedekar, Yogesh, Mahendru, Ashish, Deshmukh, Unmesh Anil
• Primary Examiner(s): Lewis, Cheryl
• Application Number:
  US14/861,687
• Publication Number:
  US 20160085810A1
• Time in Patent Office: 1,141 Days
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries

Abstract

Some event ordering requirements can be determined based on continuous event processing queries. Other event ordering requirements can be determined based on distribution flow types being used to distribute events from event streams to node executing the queries. Events from event streams can be ordered according to ordering semantics that are based on a combination of all of these event ordering requirements. Additionally, virtual computing nodes can be associated with constraints, and computing processors can be associated with capabilities. Virtual computing nodes for processing event streams can be assigned to execute on various computing processors based on both these constraints and capabilities. Additionally, for each of several events in an event stream, a ratio between a total latency and a communication latency can be for determined. Based on an average of these ratios, a quantity of reducing nodes that will be involved in a map-reduce operation can be selected.

Claims

1. A method comprising:

determining a first event ordering requirement based on a first continuous event processing query;

determining a second event ordering requirement based on a second continuous event processing query;

determining a first type of distribution flow to be used to distribute events from a first event stream to at least a first computing node of a plurality of computing nodes that are configured to execute the first continuous event processing query;

determining a second type of distribution flow to be used to distribute events from a second event stream to at least a second computing node of the plurality of computing nodes that are configured to execute the second continuous event processing query;

ordering, at the first computing node, events from the first event stream according to a first ordering semantic based at least in part on the first event ordering requirement and the first type of distribution flow, the ordering comprising;

partitioning the events of the first event stream into a first event and at least a second event for processing across the first computing node and at least one additional computing node of the plurality of computing nodes;

transmitting the first event for processing by the first computing node; and

transmitting at least the second event for processing by the additional computing node, the first event and the second event processed in accordance with the first ordering semantic; and

ordering, at the second computing node, events from the second event stream according to a second ordering semantic based at least in part on the second event ordering requirement and the second type of distribution flow, the second ordering semantic being different from the first ordering semantic.

2. The method of claim 1, wherein the plurality of computing nodes correspond to separate computing devices.

3. The method of claim 1, wherein the first type of distribution flow or the second type of distribution flow comprises at least one of a partition distribution flow, a fan-in distribution flow, a load balance distribution flow, a broadcast distribution flow or a clustered domain distribution flow.

4. The method of claim 1, wherein the first type of distribution flow or the second type of distribution flow is represented by at least one of an event processing network, a load-balancing event processing network, a broadcast event processing network, a clustered-broadcast event processing network, a partitioned event processing network or a fan-in event processing network.

5. The method of claim 1, wherein the first ordering semantic or the second ordering semantic comprises at least one of an unordered semantic, a partial-partition-ordered semantic, a total-partition-ordered semantic, a partial-ordered semantic or a total-ordered semantic.

6. The method of claim 1, further comprising:

associating the first computing node of the plurality of computing nodes with a first set of constraints indicating a first minimum quantity of memory and a first minimum quantity of processing threads;

associating the second computing node of the plurality of computing nodes with a second set of constraints indicating a second minimum quantity of memory and a second minimum quantity of processing threads;

associating a first processor of a computing device with a first set of capabilities indicating a first amount of available memory and a first thread pool size; and

associating a second processor of the computing device with a second set of capabilities indicating a second amount of available memory and a second thread pool size.

7. The method of claim 6, further comprising:

selecting, from among the plurality of computing nodes, the first computing node to be executed on the first processor based at least in part on the first set of constraints and the first set of capabilities;

updating the first set of capabilities to reflect a use by the first computing node of a portion of available memory of the first processor and a portion of a thread pool of the first processor;

selecting, from among the plurality of computing nodes, the second computing node to be executed on the second processor based at least in part on the second set of constraints and the second set of capabilities; and

updating the second set of capabilities to reflect a use by the second computing node of a portion of available memory of the second processor and a portion of a thread pool of the second processor.

8. The method of claim 1, further comprising:

for each particular event of a plurality of events in the first event stream or the second event stream;

determining a total latency for the particular event; and

determining a communication latency for the particular event.

9. The method of claim 8, further comprising:

determining a ratio of the total latency to the communication latency for the particular event;

determining an average of the ratios determined for the plurality of events; and

selecting, based on the average, a quantity of reducing nodes that will be involved in a map-reduce operation performed relative to the first event stream or the second event stream.

10. A computer-readable medium storing computer-executable instructions that, when executed by one or more processors, configures one or more computer systems to perform at least:

instructions that cause the one or more processors to determine a first event ordering requirement based on a first continuous event processing query;

instructions that cause the one or more processors to determine a second event ordering requirement based on a second continuous event processing query;

instructions that cause the one or more processors to determine a first type of distribution flow to be used to distribute events from a first event stream to at least a first computing node of a plurality of computing nodes that are configured to execute the first continuous event processing query;

instructions that cause the one or more processors to determine a second type of distribution flow to be used to distribute events from a second event stream to at least a second computing node of the plurality of computing nodes that are configured to execute the second continuous event processing query;

instructions that cause the one or more processors to order events from the first event stream according to a first ordering semantic at the first computing node based at least in part on the first event ordering requirement and the first type of distribution flow, the instructions that cause the one or more processors to order events further comprising;

instructions that cause the one or more processors to partition the events of the first event stream into a first event and at least a second event for processing across the first computing node and at least one additional computing node of the plurality of computing nodes;

instructions that cause the one or more processors to transmit the first event for processing by the first computing node; and

instructions that cause the one or more processors to transmit at least the second event for processing by the additional computing node, the first event and the second event processed in accordance with the first ordering semantic; and

instructions that cause the one or more processors to order, at the second computing node, events from the second event stream according to a second ordering semantic based at least in part on the second event ordering requirement and the second type of distribution flow, the second ordering semantic being different from the first ordering semantic.

11. The computer-readable medium of claim 10, wherein the one or more computer systems are further configured to perform instructions that cause the one or more processors to:

associate the first computing node of the plurality of computing nodes with a first set of constraints indicating a first minimum quantity of memory and a first minimum quantity of processing threads;

associate the second computing node of the plurality of computing nodes with a second set of constraints indicating a second minimum quantity of memory and a second minimum quantity of processing threads;

associate a first processor of a computing device with a first set of capabilities indicating a first amount of available memory and a first thread pool size; and

associate a second processor of the computing device with a second set of capabilities indicating a second amount of available memory and a second thread pool size.

12. The computer-readable medium of claim 11, wherein the one or more computer systems are further configured to perform instructions that cause the one or more processors to:

select, from among the plurality of computing nodes, the first computing node to be executed on the first processor based at least in part on the first set of constraints and the first set of capabilities;

update the first set of capabilities to reflect a use by the first computing node of a portion of available memory of the first processor and a portion of a thread pool of the first processor;

select, from among the plurality of computing nodes, the second computing node to be executed on the second processor based at least in part on the second set of constraints and the second set of capabilities; and

update the second set of capabilities to reflect a use by the second computing node of a portion of available memory of the second processor and a portion of a thread pool of the second processor.

13. The computer-readable medium of claim 10, wherein the one or more computer systems are further configured to perform instructions that cause the one or more processors to:

for each particular event of a plurality of events in the first event stream or the second event stream;

determine a total latency for the particular event; and

determine a communication latency for the particular event.

14. The computer-readable medium of claim 13 wherein the one or more computer systems are further configured to perform instructions that cause the one or more processors to:

determine a ratio of the total latency to the communication latency for the particular event;

determine an average of the ratios determined for the plurality of events; and

select, based on the average, a quantity of reducing nodes that will be involved in a map-reduce operation performed relative to the first event stream or the second event stream.

15. The computer-readable medium of claim 10, wherein the first ordering semantic or the second ordering semantic comprises at least one of an unordered semantic, a partial-partition-ordered semantic, a total-partition-ordered semantic, a partial-ordered semantic or a total-ordered semantic.

16. A system, comprising:

a memory storing a plurality of instructions; and

a processor configured to access the memory, the processor further configured to execute the plurality of instructions to at least;

determine a first event ordering requirement based on a first continuous event processing query;

determine a second event ordering requirement based on a second continuous event processing query;

determine a first type of distribution flow to be used to distribute events from a first event stream to at least a first computing node of a plurality of computing nodes that are configured to execute the first continuous event processing query;

determine a second type of distribution flow to be used to distribute events from a second event stream to at least a second computing node of the plurality of computing nodes that are configured to execute the second continuous event processing query;

order events from the first event stream according to a first ordering semantic at the first computing node based at least in part on the first event ordering requirement and the first type of distribution flow, the instructions to order further comprising instructions to;

partition the events of the first event stream into a first event and at least a second event for processing across the first computing node and at least one additional computing node of the plurality of computing nodes;

transmit the first event for processing by the first computing node; and

transmit at least the second event for processing by the additional computing node, the first event and the second event processed in accordance with the first ordering semantic; and

order, at the second computing node, events from the second event stream according to a second ordering semantic based at least in part on the second event ordering requirement and the second type of distribution flow, the second ordering semantic being different from the first ordering semantic.

17. The system of claim 16, wherein the plurality of computing nodes correspond to separate computing devices.

18. The system of claim 16, wherein the first type of distribution flow or the second type of distribution flow is represented by at least one of an event processing network, a load-balancing event processing network, a broadcast event processing network, a clustered-broadcast event processing network, a partitioned event processing network or a fan-in event processing network.

19. The system of claim 16, wherein the first ordering semantic or the second ordering semantic comprises at least one of an unordered semantic, a partial-partition-ordered semantic, a total-partition-ordered semantic, a partial-ordered semantic or a total-ordered semantic.

20. The system of claim 16, wherein the processor is further configured to:

determine a total latency for a particular event of a plurality of events in the first event stream or the second event stream;

determine a total latency for the particular event;

determine a communication latency for the particular event;

determine a ratio of the total latency to the communication latency for the particular event;

determine an average of the ratios determined for the plurality of events; and

select, based on the average, a quantity of reducing nodes that will be involved in a map-reduce operation performed relative to the first event stream or the second event stream.