Internode data communications in a parallel computer

• US 8,650,581 B2
• Filed: 12/11/2012
• Issued: 02/11/2014
• Est. Priority Date: 11/07/2011
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Archer, Charles J., Blocksome, Michael A., Miller, Douglas R., Parker, Jeffrey J., Ratterman, Joseph D., Smith, Brian E.
• Primary Examiner(s): Puente, Emerson
• Assistant Examiner(s): VERDI, KIMBLEANN C
• Application Number:
  US13/711,108
• Publication Number:
  US 20130125140A1
• Time in Patent Office: 427 Days
• Field of Search: 719/312, 719/314
• US Class Current: 719/314
• CPC Class Codes: G06F 9/544 Buffers; Shared memory; Pipes

Abstract

Internode data communications in a parallel computer that includes compute nodes that each include main memory and a messaging unit, the messaging unit including computer memory and coupling compute nodes for data communications, in which, for each compute node at compute node boot time: a messaging unit allocates, in the messaging unit'"'"'s computer memory, a predefined number of message buffers, each message buffer associated with a process to be initialized on the compute node; receives, prior to initialization of a particular process on the compute node, a data communications message intended for the particular process; and stores the data communications message in the message buffer associated with the particular process. Upon initialization of the particular process, the process establishes a messaging buffer in main memory of the compute node and copies the data communications message from the message buffer of the messaging unit into the message buffer of main memory.

Claims

1. A method of internode data communications in a parallel computer, the parallel computer comprising a plurality of compute nodes, each compute node comprising main computer memory and a messaging unit, the messaging unit comprising messaging unit computer memory, the messaging unit comprising a module of automated computing machinery coupling the plurality of compute nodes for data communications, the method comprising:

for each compute node at compute node boot time;

allocating, by the messaging unit in the messaging unit'"'"'s computer memory, a predefined number of message buffers, each message buffer associated with a process to be initialized on the compute node;

receiving, by the messaging unit prior to initialization of a particular process on the compute node, a data communications message intended for the particular process;

storing the data communications message in the message buffer associated with the particular process;

upon initialization of the particular process, establishing, by the particular process, a messaging buffer in the main computer memory of the compute node; and

the parallel computer comprises a parallel active messaging interface ('"'"'PAMI'"'"') and the plurality of compute nodes execute a parallel application, the PAMI comprises data communications endpoints, each data communications endpoint comprising a specification of data communications parameters for a thread of execution on one of the plurality of compute nodes, including specifications of a client, a context, and a task, the data communication endpoints coupled for data communications through the PAMI;

the particular process comprises one of the data communications endpoints;

each client comprises a collection of data communications resources dedicated to exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution;

each task represents a process of execution of the parallel application; and

copying, by the particular process, the data communications message from the message buffer of the messaging unit into the message buffer of the main computer memory, wherein each context carries out data communications operations independently and in parallel with other contexts.

2. The method of claim 1 further comprising:

maintaining, by each messaging unit, a status of unused message buffers in the messaging unit'"'"'s computer memory;

requesting, by the particular process on the compute node via a broadcast operation, availability of each of the messaging unit message buffers;

receiving, from each messaging unit, an indication of available messaging unit message buffers; and

sending, by the particular process, a data communications message to one of the available messaging unit message buffers with an instruction to initialize an on-demand process associated with the one of the available messaging unit message buffers.

3. The method of claim 1 further comprising:

receiving, by the messaging unit, a request to initialize an on-demand process to execute a task, including receiving a data communications message intended for the on-demand process;

storing, by the messaging unit in the messaging unit'"'"'s computer memory, the data communications message intended for the on-demand process in an available message buffer;

initializing the on-demand process;

establishing, by the on-demand process upon initialization, a messaging buffer in the main computer memory of the compute node;

copying, by the on-demand process, the data communications message intended for the on-demand process from the message buffer of the messaging unit into the message buffer of the main computer memory; and

upon completion of the task, releasing, by the on-demand process, the message buffer in the main computer memory and the message buffer in the messaging unit.

4. The method of claim 1 wherein each context carries out, through post and advance functions, data communications for the parallel application on data communications resources in exclusive possession of that context.