This article illustrates the mechanisms adopted in SSD write protection and compares their disadvantages and complexity.
Product reliability is the key element in its industrial applications while product quality and reliable improvements the focus of every manufacturer. SSD "write protection" technology is designed to protect data integrity. This write protection function is enabled by hardware or software settings. When SSD firmware finds that the function is enabled, it continues to accept the write-in command from the host without writing data into NAND Flash. That is, it maintains the integrity of data in SSD by setting disks into read-only and disabling the write-in command by the host.
As shown in the above figure (Figure 1), the first common practice is to set the write protection function by software: users set up host software to send the vendor command to SSD, the latter then logs the write protection to maintain its validity even after power outage and notify the FTL (Flash Translation Layer) that the write protection write function is now enabled. The FTL then discontinues any NAND write-in functions for write-in relevant commands, including TRIM and security erase, received later. This is not the case with the background write-in function of SSD including Data Scan, Refresh, and Flush Log. The merits of this are that the write protection function can be exercised by settings in software and firmware without any change in hardware.
As shown in the figure above (Figure 2), another write protection method is initiated by hardware settings. The SSD PCB features one hardware switch connecting to the GPIO pin on the SSD controller. Users turn on the switch to enable the write protection function: the SSD notifies FTL that write protection is active once the GPIO pin signal is detected. The remaining steps of the process are the same as method 1. The merits of this method are in its simplicity. What is required to enable the write protection function is turning on the hardware switch without any software settings. This is the write protection method adopted by scores of SD cards.
The third method, as shown in the figure above (Figure 3), is completely hardware based. The SSD comes with a switch connecting to the write protection pin on the NAND Flash and no changes are required on the SSD firmware. The merits of this method are that the hardware maker alone can enable the write protection function, at the price of failures of some SSD firmware, including Prevent Read Disturb, Data Retention, and Flush Log, as NAND is now totally write-in disabled.
The last write protection method is to have the function enabled by SSD firmware automatically: due to the limited service life of NAND Flash, the SSD firmware may auto enable write protection when it detects the NAND Flash usage period is about to expire or when number of bad blocks overrun predefined upper limits. This prevents users from continuing to write in critical data in the SSD leading to data damage. This write protection is auto enabled without any user intervention and available on almost every market available SSD.