System and method for data error recovery in a solid state subsystem
View Patent ↗Systems and methods are disclosed for recovering from a data access error encountered in data stripes implemented in a data redundancy scheme (e.g., RAID) in a solid state storage device. In one embodiment, the storage device holds parity data in a temporary, volatile memory such as a RAM and writes the parity data to the non-volatile memory when a full stripe's worth of new write data has been written to the non-volatile memory. In one embodiment, upon detecting that a data access error has occurred in a partially written stripe, the storage device initiates a write of the parity data for the partially written stripe to the non-volatile memory and executes a RAID recovery procedure using the newly written parity data to attempt to recover from the detected error. This approach allows for a recovery from the data access error without waiting for the full stripe to be written.
1. A solid-state storage subsystem comprising:
a non-volatile memory array;
a controller configured to implement a data redundancy configuration with a plurality of data stripes in the non-volatile memory array; and
a volatile memory for temporarily storing data to be written to the non-volatile memory array and parity data for one or more of the data stripes in the non-volatile memory array, wherein each of the plurality of data stripes is of a pre-defined stripe size and wherein the controller is configured to write parity data for a data stripe to the non-volatile memory array when data already written to the non-volatile memory for the data stripe reaches the pre-defined stripe size;
wherein the controller is further configured to:
detect an occurrence of a data access error in one of the plurality of data stripes for which parity data has not been written to the non-volatile memory array; and
in response to detecting an occurrence of a data access error, move the parity data for the data stripe causing the error from the volatile memory to the non-volatile memory array ahead of a scheduled writing of the parity data, and
attempt to correct the data access error with the newly written parity data.
2. The solid-state storage subsystem of claim 1 , wherein the writing of the parity data creates a partial data stripe that is smaller than the pre-defined stripe size.
3. The solid-state storage subsystem of claim 2 , wherein the controller is further configured to combine the partial data stripe with additional data to create an expanded stripe of the pre-defined stripe size and write new parity data for the expanded stripe.
4. The solid-state storage subsystem of claim 1 , wherein the data access error is an ECC read error.
5. The solid-state storage subsystem of claim 1 , wherein the data access error is a program error.
6. The solid-state storage subsystem of claim 1 , wherein the data redundancy configuration is a redundant array of independent disks (RAID) configuration.
7. A method of recovering from a data access error in a solid-state storage subsystem, the method comprising:
implementing a data redundancy configuration with a plurality of data stripes in a non-volatile memory array of the solid-state storage subsystem;
temporarily storing, in a volatile memory of the solid-state storage subsystem, data to be written to the non-volatile memory array and parity data associated with said data to be written;
detecting an occurrence of a data access error in one of the plurality of data stripes for which parity data has not been written to the non-volatile memory array; and
in response to detecting the occurrence of a data access error, moving the parity data for said data stripe in which the data access error occurred from the volatile memory to the non-volatile memory array to make said parity data available for attempting to correct the data access error in advance of a scheduled write for the parity data, the scheduled write corresponding to a time when the data already written to the non-volatile memory for said data stripe reaches a pre-defined stripe size.
8. The method of claim 7 , wherein each of the plurality of data stripes is of the pre-defined stripe size and wherein the writing of the parity data creates a partial data stripe that is smaller than the pre-defined stripe size.
9. The method of claim 7 , further comprising:
combining the partial data stripe with additional data to create an expanded stripe of the pre-defined stripe size; and
writing new parity data for the expanded stripe.
10. The method of claim 7 , wherein the data access error is an ECC read error.
11. The method of claim 7 , wherein the data redundancy configuration is a redundant array of independent disks (RAID) configuration.
12. A solid-state storage subsystem comprising:
a non-volatile memory array;
a controller configured to implement a data redundancy configuration with a plurality of data stripes in the non-volatile memory array; and
volatile memory for temporarily storing data to be written to the non-volatile memory array and parity data for one or more of the data stripes in the non-volatile memory array;
wherein the controller is further configured to:
detect an occurrence of a data access error in one of the plurality of data stripes for which parity data has not been written to the non-volatile memory array; and
in response to detecting an occurrence of a data access error, move the parity data for the data stripe in which the data access error occurred from the volatile memory to the non-volatile memory array, in advance of a time when the data written to the non-volatile memory for the data stripe reaches a pre-defined stripe size, whereby the data access error can be corrected with the parity data.
13. The solid-state storage subsystem of claim 12 , wherein the data access error is an ECC read error.
14. The solid-state storage subsystem of claim 12 , wherein the data access error is a program error.
15. The solid-state storage subsystem of claim 12 , wherein the data redundancy configuration is a redundant array of independent disks (RAID) configuration.