Dynamic monitoring of ability to reassemble streaming data across multiple channels based on history

• US 8,489,967 B2
• Filed: 04/03/2012
• Issued: 07/16/2013
• Est. Priority Date: 04/14/2009
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Arimilli, Ravi K., Chaudhary, Piyush
• Primary Examiner(s): Iqbal, Nadeem
• Application Number:
  US13/438,227
• Publication Number:
  US 20120191674A1
• Time in Patent Office: 469 Days
• Field of Search: 714 18- 21, 714/26, 714/27, 714/33, 714/39, 714774-776, 714/807, 714819-821
• US Class Current: 714/774
• CPC Class Codes: G06F 9/546 Message passing systems or structures, e.g. queues
• CPC Class Codes: H04L 1/1841 Resequencing
• CPC Class Codes: H04L 69/14 Multichannel or multilink protocols

Abstract

Mechanisms are provided for processing streaming data at high sustained data rates. These mechanisms receive a plurality of data elements over a plurality of non-sequential communication channels and write the plurality of data elements directly to the file system of the data processing system in an unassembled manner. The mechanisms determining whether to perform a data scrubbing operation or not based on history information indicative of whether data elements in the plurality of data elements are being received in a substantially sequential manner. The mechanisms perform a data scrubbing operation, in response to a determination to perform data scrubbing, to identify any missing data elements in the plurality of data elements written to the file system and assemble the plurality of data elements into a plurality of data streams in response to results of the data scrubbing indicating that there are no missing data elements.

Claims

1. A method, in a data processing system, for processing streaming data at high sustained data rates, comprising:

receiving a plurality of data elements, in the data processing system, over a plurality of non-sequential communication channels;

writing the plurality of data elements directly to a file system of the data processing system in an unassembled manner;

determining whether to perform a data scrubbing operation or not based on history information indicative of whether data elements in the plurality of data elements are being received in a substantially sequential manner;

performing a data scrubbing operation, in response to a determination to perform data scrubbing, to determine if there are any missing data elements that are not present in the plurality of data elements written to the file system; and

assembling the plurality of data elements into a plurality of data streams associated with the plurality of non-sequential communication channels in response to results of the data scrubbing indicating that there are no missing data elements.

2. The method of claim 1, wherein the data scrubbing operation is performed in response to a determination that the data elements in the plurality of data elements are not being received in a substantially sequential manner.

3. The method of claim 2, further comprising:

inhibiting performance of the data scrubbing operation in response to a determination that the data elements in the plurality of data elements are being received in a substantially sequential manner.

4. The method of claim 3, further comprising:

setting a flag to a first value indicating that data scrubbing is to be performed in response to a determination that the data elements in the plurality of data elements are not being received in a substantially sequential manner; and

setting the flag to a second value indicating that data scrubbing is not to be performed in response to a determination that the data elements in the plurality of data elements are being received in a substantially sequential manner, wherein if the flag is set, data scrubbing is performed and if the flag is not set, data scrubbing is not performed.

5. The method of claim 1, wherein the history information comprises data indicating an identifier of a last received data element.

6. The method of claim 2, wherein determining whether to perform data scrubbing or not comprises determining if the plurality of data elements are being received in a substantially sequential manner based on whether a current data element is within a predetermined range of data elements from a last received data element identified by the history information.

7. The method of claim 1, wherein the history information is communicated to one or more switches in a data network over which the plurality of data elements are transmitted to the data processing system, and wherein the one or more switches use the history information to select a path through the data network over which to transmit data to the data processing system.

8. The method of claim 1, further comprising:

releasing the assembled plurality of data streams for access via the file system in response to completion of the assembling of the plurality of data elements into the plurality of data streams.

9. The method of claim 1, wherein performing the data scrubbing operation comprises performing the data scrubbing operation on a metadata directory associated with the plurality of data elements to determine if the metadata indicates that metadata for portions of data are missing in the metadata directory.

10. The method of claim 1, further comprising:

in response to the data scrubbing operation indicating that there is a missing data element, transmitting a request to a source of the missing data element requesting that the missing data element be retransmitted.

11. The method of claim 1, wherein the plurality of data elements comprises portions of data elements, and wherein each portion of data elements is received from a processor in a plurality of processors.

12. The method of claim 11, wherein the plurality of processors are part of a distributed computing cluster operating on message passing interface (MPI) jobs.

13. The method of claim 12, wherein the plurality of data elements are transmitted by the plurality of processors in response to the processors encountering an MPI barrier operation.