PARALLEL COMPUTER SYSTEM AND METHOD FOR CONTROLLING PARALLEL COMPUTER SYSTEM

• US 20150341424A1
• Filed: 05/21/2015
• Published: 11/26/2015
• Est. Priority Date: 05/22/2014
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Nakashima, Makoto
• Granted Patent: US 9,942,309 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/0715 in a system implementing multitasking multitasking per se G06F9/46
• CPC Class Codes: G06F 11/0751 Error or fault detection not based on redundancy power supply failures G06F1/30; network fault management H04L41/06
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 11/302 where the computing system component is a software system
• CPC Class Codes: G06F 2201/865 Monitoring of software
• CPC Class Codes: G06F 2209/5017 Task decomposition
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• 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

Abstract

A parallel computer system includes computing nodes to execute a parallel program for generating computing processes that perform computations and monitoring processes and forming a monitoring hierarchical structure, each monitoring process monitors a monitoring process and a computing process arranged immediately lower than the monitoring process, each of the computing nodes operating as the computing process when the computing process is allocated to the computing node and operating as the monitoring process when the monitoring process is allocated to the computing node. Each of the computing nodes allocated to the monitoring process performs processing that changes the hierarchical structure based on a first target value serving as a target value for the total number of subordinate computing processes of the allocated monitoring process and a second target value serving as a target value for the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process.

Claims

1. A parallel computer system, comprising:

a plurality of computing nodes capable of executing a parallel program for generating a plurality of computing processes that perform respective predetermined computations and a plurality of monitoring processes and forming a monitoring hierarchical structure, each monitoring process monitoring a monitoring process and a computing process arranged immediately lower than the monitoring process, each of the computing nodes operating as the computing process when the computing process is allocated to the computing node and operating as the monitoring process when the monitoring process is allocated to the computing node,

each of the computing nodes, which is allocated the monitoring process, performing processing that changes the monitoring hierarchical structure based on a first target value serving as a target value for the total number of subordinate computing processes of the allocated monitoring process and a second target value serving as a target value for the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process.

2. The parallel computer system according to claim 1, wherein each of the computing nodes, which is allocated the monitoring process, controls generation and disappearance of a monitoring process immediately lower than the allocated monitoring process and rearrangement of a computing process associated with the generation and the disappearance of the immediate lower monitoring process, as the processing that changes the monitoring hierarchical structure.

3. The parallel computer system according to claim 1, wherein, when the total number of subordinate computing processes of the allocated monitoring process exceeds the first target value, each of the computing nodes, which is allocated the monitoring process, determines to change an arrangement position of an excess of a computing process over the first target value to immediately lower than a monitoring process that is a high-order process of the allocated monitoring process.

4. The parallel computer system according to claim 1, wherein, when each of the computing nodes, which is allocated the monitoring process, has a computing process arranged immediately lower than the allocated monitoring process, the number of monitoring processes and computing processes arranged immediately lower exceeds the second target value, and a subordinate monitoring process, the number of subordinate processes of which falls below the first target value, is present immediately lower than the allocated monitoring process, the computing node determines to change an arrangement position of the computing process present immediately lower than the allocated monitoring process to immediately lower than the subordinate monitoring process such that the number of subordinate processes of the subordinate monitoring process does not exceed the first target value.

5. The parallel computer system according to claim 1, wherein, when the number of monitoring processes arranged immediately lower than the allocated monitoring process falls below the second target value, and the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process surpasses the second target value, each of the computing nodes, which is allocated the monitoring process, generates a new monitoring process immediately lower than the allocated monitoring process and determines to change an arrangement position of each of a part or all of the computing processes present immediately lower than the allocated monitoring process to immediately lower than the new monitoring process.

6. The parallel computer system according to claim 1, wherein, when the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process does not exceed the second target value even after an arrangement position of a computing process present immediately lower than a given monitoring process arranged immediately lower than the allocated monitoring process is changed to immediately lower than the allocated monitoring process, each of the computing nodes, which is allocated the monitoring process, determines to terminate the given monitoring process and change the arrangement position of the computing process present immediately lower than the given monitoring process to immediately lower than the allocated monitoring process.

7. The parallel computer system according to claim 1, wherein each of the computing nodes, which is allocated the monitoring process, determines on a value obtained by dividing the first target value by the second target value supplied to the allocated monitoring process as the first target value in a monitoring process arranged immediately lower than the allocated monitoring process.

8. The parallel computer system according to claim 1, wherein, when the number of monitoring processes arranged immediately lower than the allocated monitoring process exceeds the second target value, each of the computing nodes, which is allocated the monitoring process, determines to terminate some of the monitoring processes arranged immediately lower such that the number of monitoring processes arranged immediately lower is not more than the second target value.

9. A computer-readable recording medium having stored therein a program for causing a plurality of computing nodes to execute a process comprising:

generating a plurality of computing processes that perform respective predetermined computations and a plurality of monitoring processes; and

forming a monitoring hierarchical structure, each monitoring process monitoring a monitoring process and a computing process arranged immediately lower than the monitoring process,

the program causing each of the computing nodes to execute processing including;

operating as the computing process when the computing process is allocated; and

operating as the monitoring process when the monitoring process is allocated and changing the monitoring hierarchical structure based on a first target value serving as a target value for the total number of subordinate computing processes of the allocated monitoring process and a second target value serving as a target value for the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process.

10. A method of controlling a parallel computer system including a plurality of computing nodes capable of executing a parallel program to be executed by a plurality of computing nodes for generating a plurality of computing processes that perform respective predetermined computations and a plurality of monitoring processes and forming a monitoring hierarchical structure, each monitoring process monitoring a monitoring process and a computing process arranged immediately lower than the monitoring process, the method comprising:

each of the computing nodes operating as the computing process when the computing process is allocated, and

each of the computing nodes operating as the monitoring process when the monitoring process is allocated and performing processing that changes the monitoring hierarchical structure based on a first target value serving as a target value for the total number of subordinate computing processes of the allocated monitoring process and a second target value serving as a target value for the number of monitoring processes and computing processes arranged immediately lower than the allocated monitoring process.