It turned out that under certain customer environments network conditions and newer versions of SSD Boards with
high density flash memory, an increase of failing SSD Boards could be identified. After comprehensive
investigation of defective SSD Boards two direct causes have been found.
1. The data management table in SSD Board control memory block is lost or corrupted
2. The complete SSD Board control memory block is abnormal already from manufacturing
Number 2
was caused by a lack of inspection at the manufacturer and has been already corrected at the
manufacturer side wherefore the problem rate related to this issue, which is the
, should go
nearly to zero or will only happen in rare cases.
Number 1
happens more often and causes the "continuous rebooting problem". This problem has been solved
by new SSD Board firmware.
The cause of the problem is related in the number of read and writes to the same page within a memory block and
a firmware failure in the wear leveling process. This process normally prevents that too many read and writes can
happen on the same page/block to secure data corruption due to overuse of a memory block. While this firmware
failure does only happen within the power on/off process of the SSD Board, identification was very difficult. Even
more not only the power on/off of the SSD Board was involved but also the customer network environment, the
MFP sleep mode (in this case the deep sleep mode) and the frequency the MFP does awake from deep sleep
mode.
If the MFP is in
deep sleep mode and a network broadcast like an arp request or ping reaches the MFP it will
shortly awake from deep sleep mode and will immediately, after the request has been fulfilled, go back to deep
sleep mode. This in turn will also shortly switch on and off the SSD Board, but will not be recognized by the SSD
Board firmware wear leveling process and therefore will write always to the same page. This will happen as long
the MFP is in this condition and will not wake up completely.
Based on the above we can say that in network environments with many broadcasts, s with low usage
conditions and therefore long deep sleep mode times, the risk is high, that the data management table in SSD
Board control memory block is lost or corrupted and the "continuous rebooting problem" happens more often.
Bookmarks