Browse Prior Art Database

Force Checkpoint

IP.com Disclosure Number: IPCOM000047717D
Original Publication Date: 1983-Dec-01
Included in the Prior Art Database: 2005-Feb-08
Document File: 1 page(s) / 11K

Publishing Venue

IBM

Related People

McGilvray, BL: AUTHOR [+2]

Abstract

This mechanism is used for a processor that has a checkpoint retry which encompasses multiple instructions. Conditionally, either manually or after the detection of an error, the time duration between checkpoints may be shortened by forcing the checkpoint to occur. The manual method may be used to troubleshoot a problem. The automated method, after error detection occurs, shortens the time duration between checkpoints so that a subsequent error will occur as close to a checkpoint as possible. This improves control over the status of the system when errors occur by keeping the previous checkpoint as close as possible to the point of error. This facilitates error data recording and analysis. After a limited time period (e.g., 10 ms) if no error has occurred, then the normal rate of checkpointing will be reestablished.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

Force Checkpoint

This mechanism is used for a processor that has a checkpoint retry which encompasses multiple instructions. Conditionally, either manually or after the detection of an error, the time duration between checkpoints may be shortened by forcing the checkpoint to occur. The manual method may be used to troubleshoot a problem. The automated method, after error detection occurs, shortens the time duration between checkpoints so that a subsequent error will occur as close to a checkpoint as possible. This improves control over the status of the system when errors occur by keeping the previous checkpoint as close as possible to the point of error. This facilitates error data recording and analysis. After a limited time period (e.g., 10 ms) if no error has occurred, then the normal rate of checkpointing will be reestablished. Thus, performance is not degraded by more frequent checkpoints, except during the time period immediately after an error is detected.

1