System and method for parallel multiplexing between servers in a cluster

• US 8,756,329 B2
• Filed: 05/17/2011
• Issued: 06/17/2014
• Est. Priority Date: 09/15/2010
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Reynolds, Nathan, Otenko, Oleksandr, Dorr, Joshua
• Primary Examiner(s): Haile, Awet
• Assistant Examiner(s): MYERS, ERIC A
• Application Number:
  US13/109,871
• Publication Number:
  US 20120066400A1
• Time in Patent Office: 1,127 Days
• Field of Search: 709223-228, 709230-234, 709/242
• US Class Current: 709/228
• CPC Class Codes: G06F 15/17318 Parallel communications techniques, e.g. gather, scatter, reduce, roadcast, multicast, all to all
• CPC Class Codes: G06F 16/20 of structured data, e.g. relational data
• CPC Class Codes: G06F 16/2358 Change logging, detection, and notification replication G06F16/27
• CPC Class Codes: G06F 16/252 between a Database Management System and a front-end application
• CPC Class Codes: G06F 16/278 Data partitioning, e.g. horizontal or vertical partitioning
• CPC Class Codes: G06F 9/45504 Abstract machines for programme code execution, e.g. Java virtual machine [JVM], interpreters, emulators
• CPC Class Codes: H04L 41/042 comprising distributed management centres cooperatively managing the network
• CPC Class Codes: H04L 41/0686 Additional information in the notification, e.g. enhancement of specific meta-data
• CPC Class Codes: H04L 49/358 Infiniband Switches
• 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: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1004 Server selection for load balancing
• CPC Class Codes: H04L 67/1027 Persistence of sessions during load balancing
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
• CPC Class Codes: H04L 67/14 Session management for real-time applications in data packet communications networks H04L65/1066
• CPC Class Codes: H04L 67/142 Managing session states for stateless protocols; Signalling session states; State transitions; Keeping-state mechanisms
• CPC Class Codes: H04L 67/146 Markers for unambiguous identification of a particular session, e.g. session cookie or URL-encoding
• CPC Class Codes: H04L 69/14 Multichannel or multilink protocols
• CPC Class Codes: Y10T 409/308904 : Multiple row dividing head
• CPC Class Codes: Y10T 409/309072 : including cutter limited to rotary motion
• CPC Class Codes: Y10T 409/309128 : with means to adjust work support vertically
• CPC Class Codes: Y10T 409/309352 : Cutter spindle or spindle support
• CPC Class Codes: Y10T 409/309408 : with cutter holder

Abstract

Systems and methods are provided for providing parallel muxing between servers in a cluster. One such system can include a cluster of one or more high performance computing systems, each including one or more processors and a high performance memory. The cluster communicates over an InfiniBand network. The system can also include a middleware environment, executing on the cluster, that includes one or more application server instances. The system can further include a plurality of muxers, wherein each application server instance includes at least one muxer. Each muxer can receive information from a plurality of threads to transmit to a different muxer on a different application server instance over the Infiniband network using a plurality of parallel channels.

Claims

1. A system for providing parallel multiplexing between servers in a cluster, comprising:

a cluster of one or more high performance computing systems, each including one or more processors and a high performance memory, and wherein the cluster communicates over an InfiniBand network;

a middleware environment, executing on the cluster, that includes a plurality of application server instances;

a plurality of multiplexers, wherein each of the plurality of multiplexers resides on a said application server instance and is directly connected with a plurality of parallel channels, wherein each of the plurality of parallel channels is

created using a replication channel configured for the cluster as a template and overwriting port information therein with a unique port number, and

further connected to a different multiplexer on a different application server instance;

a sender queue directly connected with each said multiplexier, wherein the sender queue aggregates messages received from a plurality of threads and places the messages sequentially therein; and

a plurality of sender threads, wherein each of said sender threads operates on one of said plurality of parallel channels;

wherein each said multiplexer is operable to

flush the messages from the sender queue,

distribute the flushed messages over the plurality of parallel channels, and

use the plurality of sender threads to transmit the flushed messages concurrently over the plurality of parallel channels to the different multiplexer on the different application server instance over the Infiniband network.

2. The system of claim 1, wherein a user can configure how many parallel channels are included in the plurality of parallel channels.

3. The system of claim 1, wherein the messages include session data.

4. The system of claim 1, wherein each said multiplexer is a New I/O (NIO) multiplexer.

5. The system of claim 1, further comprising a utility function used to determine if multiple channels are configured in response to configuration changes and cache the result for subsequent calls.

6. A method for providing parallel multiplexing between servers in a cluster, comprising:

providing a cluster of one or more high performance computing systems, each including one or more processors and a high performance memory, and wherein the cluster communicates over an InfiniBand network;

providing a middleware environment, executing on the cluster, that includes a plurality of application server instances;

providing a plurality of multiplexers, wherein each said mutiplexier resides on a said application server instance, and is directly connected with a plurality of parallel channels, wherein each of the plurality of parallel channels is

created using a replication channel configured for the cluster as a template and overwriting port information therein with a unique port number, and

further connected to a different multiplexier at a different application server instance;

providing a sender queue directly connected with each said multiplexier, wherein the sender queue aggregates messages received from a plurality of threads and places the messages sequentially therein; and

providing a plurality of sender threads, wherein each of said sender threads operates on one of said plurality of parallel channels;

wherein each said multiplexer is operable to

flush the messages from the sender queue,

distribute the flushed messages over the plurality of parallel channels, and

use the plurality of sender threads to transmit the flushed messages concurrently over the plurality of parallel channels.

7. The method of claim 6, further comprising configuring how many parallel channels are included in the plurality of parallel channels, based on input from a user.

8. The method of claim 6, wherein the messages include session data.

9. The method of claim 6, wherein each said multiplexer is a New I/O (NIO) multiplexer.

10. The method of claim 6, wherein a utility function is used to determine if multiple channels are configured in response to configuration changes and cache the result for subsequent calls.

11. A non-transitory computer readable storage medium, including instructions stored there on which, when executed by a computer, cause the computer to perform the steps comprising:

providing a cluster of one or more high performance computing systems, each including one or more processors and a high performance memory, and wherein the cluster communicates over an InfiniBand network;

providing a middleware environment, executing on the cluster, that includes a plurality of application server instances;

providing a plurality of multiplexers, wherein each said mutiplexier resides on a said application server instance, and is directly connected with a plurality of parallel channels, wherein each of the plurality of parallel channels is

created using a replication channel configured for the cluster as a template and overwriting port information therein with a unique port number, and

further connected to a different multiplexer at a different application server instance;

providing a sender queue directly connected with each said multiplexier, wherein the sender queue aggregates messages received from a plurality of threads and sequentially place the messages therein; and

providing a plurality of sender threads, wherein each of said sender threads operates on one of said plurality of parallel channels;

wherein each said multiplexer is operable to

flush the messages from the sender queue,

distribute the flushed messages over the plurality of parallel channels, and

use the plurality of sender threads to transmit the flushed messages concurrently over the plurality of parallel channels.

12. The non-transitory computer readable storage medium of claim 11, further comprising configuring how many parallel channels are included in the plurality of parallel channels, based on input from a user.

13. The non-transitory computer readable storage medium of claim 11, wherein the messages include session data.

14. The non-transitory computer readable storage medium of claim 11, wherein each said multiplexer is a New I/O (NIO) multiplexer.

15. The non-transitory computer readable storage medium of claim 11, wherein a utility function is used to determine if multiple channels are configured in response configuration changes and cache the result for subsequent calls.