Transparent and lightweight recovery from hardware memory errors

摘要

Systems and methods are disclosed that allow for transparently recovering from an uncorrected multi-bit error of arbitrary length located at a memory address. Storing one or more parity pages, for a set of pages in system memory, such that a page in the set of pages may be reconstructed using one of the parity pages is disclosed. Storing an indication of one or more page'disk location such that the one or more pages may be reconstructed by refilling the page from disk is also disclosed.

主权项

1. A system comprising:
a memory for storing a plurality of pages and a second plurality of pages; and a processor programmed to: store an indication of a disk location of each of the plurality of pages such that each of the plurality of pages may be reconstructed by refilling the page from disk in response to a memory error on the page, wherein the error in the memory is a corrupted page of the plurality of pages; and store parity pages such that any page in the second plurality of pages may be reconstructed using one of the parity pages.