Method for distributed RDSMS

• US 8,521,770 B1
• Filed: 05/15/2012
• Issued: 08/27/2013
• Est. Priority Date: 03/15/2003
• Status: Expired due to Term

• Current Assignee: Guavus, Inc. (Thales SA)
• Original Assignee: SQLstream, Inc. (Thales SA)
• Inventors: Black, Damian
• Primary Examiner(s): Fleurantin, Jean B
• Assistant Examiner(s): Ly, Anh
• Application Number:
  US13/472,259
• Time in Patent Office: 469 Days
• Field of Search: 707/713, 707/769, 707/770, 707/802, 707/805, 709/231, 709/237
• US Class Current: 707/769
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries
• CPC Class Codes: G06F 16/83 Querying
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: Y10S 707/922 Communications
• CPC Class Codes: Y10S 707/954 Relational
• CPC Class Codes: Y10S 707/959 Network
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

Distributed Data Stream Management System (DDSMS) and the optimization strategies and capabilities enabled therein, and embodiments of its stream processing nodes are provided. Data recovery both from local buffered data and through recursive recovery are possible.

Claims

1. A method of brokering messages comprising data records between publishers and subscribers using at least one data processing device, said method comprising:

said at least one data processing device comprising one or more processors;

providing at least one data stream processing program executing on said at least one data processing device for processing streams of data records;

processing said streams of data records within said data stream processing programs according to a set of relational stream processing operations that comprise stream processing queries, wherein each said at least one data processing device performs said processing without a need to include a persistent relational database element to process said streams of data records;

communicating with a collection of external data publishers to create input data streams for processing by said data stream processing programs;

communicating with a collection of external data subscribers to consume output data streams of said data stream processing programs;

allowing said external data subscribers to create subscriptions in the form of subscription queries comprising stream processing queries including relational stream processing queries;

evaluating said subscription queries against said streams by executing the stream processing operations of said subscription queries within said data stream processing programs;

generating a stream of query subscription answer records per subscription from said subscription queries;

sending said query subscription answer records for each subscription to said external data subscribers via a stream answer mechanism such that said data records can be read individually;

allowing said external data subscribers to be notified when there are at least one of said data records available for reading.

2. The method of claim 1, wherein said stream answer mechanism allows external data subscribers each to obtain a copy of results of relational subscriptions whereby said external data publishers and said external data subscribers are able to broker records.

3. The method of claim 1, wherein said stream answer mechanism maintains one or more queues of said answer records per subscription so that external data subscribers individually consume said answer records by removing said answer records from said queues such that each of said answer records is distributed to at most one of said external data subscribers whereby said external data publishers and said external data subscribers are able to broker records.

4. The method of claim 1, wherein multiple queues of varying priority are maintained, said method further comprising prioritizing accessibility by specific subscribers to said queues.

10. The method of claim 2, wherein said ensuring comprises recognizing identities of each tuple of said retransmitted records, and ignoring said retransmitted records that have already been processed.

5. A method of brokering messages comprising data records between publishers and subscribers using at least one data processing device, said method comprising:

said at least one data processing device comprising one or more processors;

providing at least one data stream processing program executing on said at least one data processing device for processing streams of data records;

processing said streams of data records within said data stream processing programs according to a set of relational stream processing operations that comprise stream processing queries, wherein each said at least one data processing device performs said processing without a need to include a relational database element to process said streams of data records;

wherein multiple, differing views of said data streams are defined;

communicating with a collection of external data publishers to create input data streams for processing by said data stream processing programs; and

communicating with a collection of external data subscribers to consume output data streams of said data stream processing programs;

wherein said data stream processing programs are written that process said data records sequentially.

6. The method of claim 5, wherein said method provides distributed, peer-to-peer applications between said publishers and said subscribers.

7. The method of claim 5, wherein said processing programs include structured loops configured to process records in said streams as they arrive at said at least one data processing device, respectively.

8. The method of claim 5, wherein recursive recovery is provided to prevent loss of records from a data stream received at said at least one data processing device, respectively.

9. A method of managing and processing streams of data records over a data communication network linking a plurality of data processors, said method comprising:

said at least one data processors comprising one or more processors;

providing a plurality of stream processing nodes for consuming and generating streams of data records executing relational streaming operations on said data processors, wherein said stream processing nodes perform said consuming and generating without a need to include a persistent relational database element to consume and generate said streams of data records;

receiving transmission of said streams of data records by said stream processing nodes;

buffering outputs of said streams of data records by said stream processing nodes;

retransmitting a least a portion of said outputs buffered by at least one of said stream processing nodes when requested by at least one other of said stream processing nodes;

disposing of said outputs having been buffered when it is confirmed that all of said stream processing nodes programmed to receive said outputs have received said outputs; and

further comprising ensuring that no retransmitted outputs that have been previously processed by any of said stream processing nodes are processed again by any of the same said stream processing nodes.