Providing parity in a RAID sub-system using non-volatile memory

A method for providing parity correction for a RAID array in a computer system after a system failure, including steps of:

• maintaining information about stripes having possibly inconsistent parity caused by updating data during routine operation of the computer system;
• after said system failure, identifying stripes in response to said information;
• and
• for each said stripe so identified, correcting said possibly inconsistent parity;
• wherein said information includes a bitmap and including the further step of, when said bitmap is sufficiently full and after all pending writes are successfully completed, clearing said bitmap;
• wherein each bit in said bitmap is associated with a group of one or more stripes;
• and
• wherein said step of maintaining includes steps of:
• when updating data, a portion of said data destined for a first one or more blocks of a stripe having an associated group and an existing parity, reading a second one or more blocks of said stripe used to compute a new parity of said stripe;
• computing said new parity for said stripe;
• selecting a bit in said bitmap responsive to said group;
• setting said bit;
• writing said portion of said data to said stripe;
• writing said new parity to said stripe.

An apparatus for providing parity correction for a RAID array in a computer system after a system failure, said apparatus including:

• a maintenance mechanism configured to maintain information about stripes having possibly inconsistent parity caused by updating data during routine operation of the computer system;
• an identification mechanism configured to identify stripes in response to said information, said identification apparatus activated in response to check on reboot after a system failure;
• a correction mechanism configured to correct said possibly inconsistent parity for each of said stripes identified by said identification mechanism;
• wherein said maintenance mechanism further includes:
• a read mechanism configured to read a second one or more blocks of said stripe required to compute a new parity of said stripe, said mechanism activated when new data is to be written to a first one or more blocks of a stripe having a stripe number and an existing parity;
• a compute mechanism configured to compute said new parity for said stripe responsive to said reading;
• an add mechanism configured to add said stripe number to a list of dirty stripes responsive to said compute mechanism;
• a write mechanism configured to write said data to said stripe;
• a second write mechanism configured to write said new parity to said stripe;
• a remove mechanism configured to remove said stripe number from said list of dirty stripes.

An apparatus for providing parity correction for a RAID array in a computer system after a system failure, said apparatus including:

• a maintenance mechanism configured to maintain information about stripes having possibly inconsistent parity caused by updating data during routine operation of the computer system;
• an identification mechanism configured to identify stripes in response to said information, said identification apparatus activated in response to check on reboot after a system failure;
• a correction mechanism configured to correct said possibly inconsistent parity for each of said stripes identified by said identification mechanism;
• wherein said information includes a bitmap and further including a clear mechanism configured to clear said bitmap when said bitmap is sufficiently full and after all pending writes are successfully completed;
• and
• wherein said maintenance mechanism includes:
• a read mechanism configured to read a first one or more blocks of said stripe used to compute a new parity of a stripe, a second one or more blocks of said stripe to be updated with new data, said stripe having an associated group and an associated parity block;
• a compute mechanism configured to compute said new parity for said stripe, responsive to said read mechanism;
• a selector configured to select a bit in said bitmap responsive to said group;
• a set mechanism configured to set said bit;
• a first write mechanism configured to write new data to said stripe;
• a second write mechanism configured to write said new parity to said stripe.

A memory storing information including instructions, the instructions executable by a processor to provide parity correction for a RAID array in a computer system after a system failure, the instructions including steps of:

• maintaining information about stripes having possibly inconsistent parity during routine operation of the computer system;
• after said system failure, identifying stripes in response to said information;
• and
• for each said stripe so identified, correcting said possibly inconsistent parity;
• wherein said step of maintaining includes steps of:
• when updating data, a portion of said data destined for a first one or more blocks of a stripe having a stripe number and an existing parity, reading a second one or more blocks of said stripe required to compute a new parity of said stripe;
• computing said new parity for said stripe;
• adding said stripe number to a list of dirty stripes;
• writing said portion of said data to said stripe;
• writing said new parity to said stripe;
• removing said stripe number from said list of dirty stripes.

A memory storing information including instructions, the instructions executable by a processor to provide parity correction for a RAID array in a computer system after a system failure, the instructions including steps of:

• maintaining information about stripes having possibly inconsistent parity caused by updating data during routine operation of the computer system;
• after said system failure, identifying stripes in response to said information;
• and
• for each said stripe so identified, correcting said possibly inconsistent parity;
• wherein said information includes a bitmap and including the further step of, when said bitmap is sufficiently full and after all pending writes are successfully completed, clean said bitmap;
• wherein each bit in said bitmap is associated with a group of one or more stripes;
• and
• wherein said step of maintaining includes steps of:
• when updating data, a portion of said data destined for a first one or more blocks of a stripe having an associated group and an existing parity, reading a second one or more blocks of said stripe used to compute a new parity of said stripe;
• computing said new parity for said stripe;
• selecting a bit in said bitmap responsive to said group;
• setting said bit;
• writing said portion of said data to said stripe;
• writing said new parity to said stripe.