System and method for preventing solid state drive corruption after dirty shutdown power loss

Patent metadata

Field	Value
Publication number	US-9891846-B2
Application number	US-201514675299-A
Country	US
Kind code	B2
Filing date	Mar 31, 2015
Priority date	Mar 31, 2015
Publication date	Feb 13, 2018
Grant date	Feb 13, 2018

Abstract

An information handling system a first controller and a solid state drive. The first controller receives an indication that a forced shutdown of the information handling system has been initiated, and triggers a cache flush command in response to receiving the indication that the forced shutdown has been initiated. The solid state drive includes a memory to store data, a cache to buffer data prior to writing the data in the memory, and a second controller. The second controller receives the cache flush command, and flushes the cache by writing the data in the cache to the memory in response to the cache flush command being received.

First claim

What is claimed is:

1. An information handling system comprising:
a first controller to receive an indication that a forced shutdown of the information handling system has been initiated, and to trigger a cache flush command in response to receiving the indication that the forced shutdown has been initiated; and
a solid state drive configured to communicate with the first controller, the solid state drive including:
a memory to store data;
a cache to buffer data prior to writing the data in the memory; and
a second controller configured to communicate with the cache and with the memory, the second controller to write the data in the cache to the memory in response to a pre-determined amount of data being stored in the cache, the second controller to receive the cache flush command prior to the pre-determined amount of data being stored in the cache, in response to the cache flush command being received, to flush the cache by writing the data in the cache to the memory prior to the forced shutdown being executed, wherein the data flushed from the cache is written to the memory via a wear leveling write, wherein accesses to different portions of the memory are even during the wear leveling write, the second controller to resume normal operations after the flush of the cache and in response to the forced shutdown being aborted, wherein the flush of the cache results in an additional wear leveling write to the memory based on the data in the cache being stored in the memory prior to the pre-determined amount of data being stored in the cache and the forced shutdown being aborted.

2. The information handling system of claim 1 , wherein the forced shutdown has been initiated for a predetermined amount of time before the forced shutdown is executed.

3. The information handling system of claim 1 , further comprising:
a dedicated connector between the first controller and the second controller, wherein the first controller provides the cache flush command to the second controller over the dedicated connector, wherein the dedicated connector is a physical connector located only between the first controller and the second controller.

4. The information handling system of claim 3 , wherein the cache flush command being provided over the dedicated connector results in the flush of the cache without causing a break in a SATA command sequence.

5. The information handling system of claim 1 , further comprising:
a basic input/output system configured to communicate with the controller and with the second controller, the basic input/output system to receive a system management interrupt from the first controller, and to provide the cache flush command to the second controller in response to the system management interrupt.

6. The information handling system of claim 5 , wherein the basic input/output system provides the cache flush command to the second controller via a standby immediately command, wherein the standby immediately command causes the second controller to flush the cache by writing the data in the cache to the memory, and causes a break in a SATA command sequence.

7. A method comprising:
buffering data in a cache of a solid state drive prior to writing the data to a memory of the solid state drive, wherein the data in the cache is written to the memory in response to a pre-determined amount of data being stored in the cache;
detecting, at a first controller, an indication that a forced shutdown of an information handling system has been initiated;
triggering, by the first controller, a cache flush command in response to receiving the indication that the forced shutdown has been initiated;
receiving, at a second controller, the cache flush command prior to the pre-determined amount of data being stored in the cache;
in response to receiving the cache flush command, writing the data from the cache of the solid state drive to the memory of the solid state drive to flush the cache prior to the forced shutdown being executed, wherein the data flushed from the cache is written to the memory via a wear leveling write, wherein accesses to different portions of the memory are even during the wear leveling write; and
resuming, at the second controller, normal operations after the cache is flushed and in response to the forced shutdown being aborted, wherein the flush of the cache results in an additional wear leveling write to the memory based on the data in the cache being stored in the memory prior to the pre-determined amount of data being stored in the cache and the forced shutdown being aborted.

8. The method of claim 7 , wherein the forced shutdown has been initiated for a predetermined amount of time before the forced shutdown is executed.

9. The method of claim 7 , wherein triggering the cache flush command comprises:
providing, by the first controller, the cache flush command to the second controller over a dedicated connector, wherein the dedicated connector is a physical connector located only between the first controller and the second controller.

10. The method of claim 9 , wherein providing the cache flush command over the dedicated connector results in the flush of the cache without causing a break in a SATA command sequence.

11. The method of claim 7 , wherein triggering the cache flush command comprises:
providing, by the first controller, a system management interrupt to a basic input/output system;
receiving, at the basic input/output system, the system management interrupt; and
providing, by the basic input/output system, a cache flush command to the second controller in response to the system management interrupt.

12. The method of claim 11 , wherein the basic input/output system provides the cache flush command to the second controller via a standby immediately command, wherein the standby immediately command causes the second controller to flush the cache by writing the data in the cache to the memory, and causes a break in a SATA command sequence.

13. A non-transitory computer readable medium when executed by a processor to cause the processor to implement a method, the method comprising:
buffering data in a cache of a solid state drive prior to writing the data to a memory of the solid state drive, wherein the data in the cache is written to the memory in response to a pre-determined amount of data being stored in the cache;
detecting an indication that a forced shutdown of an information handling system has been initiated;
triggering a cache flush command in response to receiving the indication that the forced shutdown has been initiated;
receiving the cache flush command prior to the pre-determined amount of data being stored in the cache;
in response to receiving the cache flush command, writing the data from the cache of the solid state drive to the memory of the solid state drive to flush the cache prior to the forced shutdown being executed, wherein the data flushed from the cache is written to the memory via a wear leveling write, wherein accesses to different portions of the memory are even during the wear leveling write; and
resuming, at the second controller, normal operations after the cache is flushed and in response to the forced shutdown being aborted, wherein the flush of the cache results in an additional wear leveling write to the memory based on the data in the cache being stored in the memory prior to the pre-determined amount of data being stored in the cache and the forced shutdown being aborted.

14. The non-transitory computer readable medium of claim 13 , wherein the forced shutdown has been initiated for a predetermined amount of time before the forced shutdown is executed.

Assignees

• Dell Products Lp

Inventors

• Hsu Isaac
• Montero Adolfo S

Classifications

• G06F12/0891 Primary

  using clearing, invalidating or resetting means · CPC title

• G06F3/0679

  Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP] · CPC title

• G06F2212/313

  In storage device · CPC title

• G06F3/0659

  Command handling arrangements, e.g. command buffers, queues, command scheduling · CPC title

• G06F2212/1032

  Reliability improvement, data loss prevention, degraded operation etc · CPC title