Data retention (DR) warning system
Aspects of a storage device including a memory and a controller are provided. The controller may determine to perform garbage collection on a superblock. During the garbage collection process, the controller will typically move the superblock into an erase pool for erasing the superblock. However, aspects of the disclosure are directed to a method of measuring a raw bit error rate (RBER) of the superblock prior to erasure. The measured RBER may be used to estimate a data retention time of the storage device and provide the customer with an early warning notification if a health metric of the storage devices reaches a threshold retention time.
1. A storage device, comprising:
a memory comprising a first superblock and a second superblock; and
a controller coupled to the memory and configured to:
select the first superblock for an audit, wherein the first superblock is selected from an active pool of superblocks if no superblocks are held in an inactive pool of superblocks, and wherein the first superblock is selected from the active pool of superblocks based on the first superblock having a lowest health metric relative to other superblocks from the active pool;
audit the first superblock to determine whether an amount of invalid data contained in the first superblock satisfies a threshold condition;
determine to perform garbage collection on the first superblock based on satisfaction of the threshold condition; and
in response to the determination to perform garbage collection and prior to erasure of invalid data from the first superblock:
copy valid data from the first superblock to the second superblock if the first superblock is selected from the active pool of superblocks; and
measure a first raw bit error rate (RBER) of the first superblock; and
estimate a data retention time of the storage device based on the first RBER and a second RBER measured from the second superblock prior to the first superblock measurement.
2. The storage device of claim 1 , wherein the controller is further configured to measure the first RBER from one or more of an oldest wordline of the first superblock or a weakest wordline of the first superblock.
3. The storage device of claim 1 , wherein the controller is further configured to erase the first superblock after measuring the first RBER of the first superblock.
4. The storage device of claim 1 , wherein the estimated data retention time is further based on a threshold RBER indicative of a configured maximum bit error rate.
5. The storage device of claim 4 , wherein the controller is further configured to output a warning notification to a user of the storage device if the estimated data retention time is equal to a current time.
6. The storage device of claim 1 , wherein the controller, being configured to audit the first superblock, is further configured to select the first superblock for audit based on a scheduled event.
7. The storage device of claim 1 , wherein the controller is further configured to store the estimated data retention time in a self-monitoring, analysis, and reporting technology (SMART) log.
8. A method for self-testing by a storage device, the method comprising:
selecting a first superblock for an audit, wherein the first superblock is selected from an active pool of superblocks if no superblocks are held in an inactive pool of superblocks, and wherein the first superblock is selected from the active pool of superblocks based on the first superblock having a lowest health metric relative to other superblocks from the active pool;
auditing the first superblock of the storage device to determine whether an amount of invalid data contained in the first superblock satisfies a threshold condition;
determining to perform garbage collection on the first superblock based on satisfaction of the threshold condition; and
in response to the determination to perform garbage collection and prior to erasure of invalid data from the first superblock:
copying valid data from the first superblock to a second superblock of the storage device if the first superblock is selected from the active pool of superblocks; and
measuring a first raw bit error rate (RBER) of the first superblock; and
estimating a data retention time of the storage device based on the first RBER and a second RBER measured from the second superblock prior to the first superblock measurement.
9. The method of claim 8 , further comprising measuring the first RBER from one or more of an oldest wordline of the first superblock or a weakest wordline of the first superblock.
10. The method of claim 8 , further comprising erasing the first superblock after measuring the first RBER of the first superblock.
11. The method of claim 10 , wherein the estimated data retention time is further based on a threshold RBER indicative of a configured maximum bit error rate.
12. The method of claim 8 , further comprising outputting a warning notification to a user of the storage device if the estimated data retention time is equal to a current time.
13. The method of claim 8 , further comprising selecting the first superblock for audit based on a scheduled event.
14. The method of claim 8 , further comprising storing the estimated data retention time in a self-monitoring, analysis, and reporting technology (SMART) log.