> Note There is no counter or other attribute that reports drive idle power-on hours. Intel recommends that you use SMART 09h (drive power-on hours) as an approximation of the idle power-on hours. Therefore, Intel strongly recommends that you update the firmware as soon as possible to avoid the risk of unrecoverable data loss. The new MR1 firmware will not resolve any read-uncorrectable errors that existed before the firmware upgrade.
> Intel S4510 or S4610 SATA SSD (only 1.8 TB and 3.84 TB are affected) can experience the following issues after 1700+ cumulative idle hours.
> Users can see critical errors under certain use conditions. The drive may report BAD_CONTEXT_2033, BAD_CONTEXT_1042 or excessive LBA mismatch (SMART attribute B8h) after 1700 hours of cumulative power-on idle time.
> Once the error code is triggered, the drive will fail to respond on the bus.
I bet spinning platter hard drives could suffer a similar firmware failure, no?
Would the technical challenge of changing logic boards to access platters be more easily remedied, than say, reflashing the firmware of an SSD?
If so, really, that precise concern is a deeply flawed aspect of the design. Wear leveling and block indices or allocation tables should have a boundary layer isolating them from your data to some degree.
The chips that store your data should be recoverable without a firmware dependency. This is an aspect of SSD design that should receive much, if not the most attention, above and beyond capacity and speed. After all, persistent storage is fundamentally about non-volatile durability in the face of adverse conditions.
6 comments
[ 4.3 ms ] story [ 42.7 ms ] threadMicrosoft: https://support.microsoft.com/en-us/help/4499612/intel-ssd-d...
> Note There is no counter or other attribute that reports drive idle power-on hours. Intel recommends that you use SMART 09h (drive power-on hours) as an approximation of the idle power-on hours. Therefore, Intel strongly recommends that you update the firmware as soon as possible to avoid the risk of unrecoverable data loss. The new MR1 firmware will not resolve any read-uncorrectable errors that existed before the firmware upgrade.
Dell: https://www.dell.com/support/home/us/en/ilbsdt1/drivers/driv...
Lenovo: https://support.lenovo.com/us/en/solutions/ht507987
> Intel S4510 or S4610 SATA SSD (only 1.8 TB and 3.84 TB are affected) can experience the following issues after 1700+ cumulative idle hours.
> Users can see critical errors under certain use conditions. The drive may report BAD_CONTEXT_2033, BAD_CONTEXT_1042 or excessive LBA mismatch (SMART attribute B8h) after 1700 hours of cumulative power-on idle time.
> Once the error code is triggered, the drive will fail to respond on the bus.
Famous Intels graceful degradation (revert to read-only)!
Would the technical challenge of changing logic boards to access platters be more easily remedied, than say, reflashing the firmware of an SSD?
If so, really, that precise concern is a deeply flawed aspect of the design. Wear leveling and block indices or allocation tables should have a boundary layer isolating them from your data to some degree.
The chips that store your data should be recoverable without a firmware dependency. This is an aspect of SSD design that should receive much, if not the most attention, above and beyond capacity and speed. After all, persistent storage is fundamentally about non-volatile durability in the face of adverse conditions.