> My concern with using it to accelerate non-volatile storage > is partly with the prospects for human error (adapter cord > to trip over, command that needs to be run), but also
The DDRdrive X1 ACDC power cord exits the rear of the chassis just as the host power supply cord, if a customer has issues with power cords being randomly removed, the root cause needs to be addressed and would not be unique to the X1. We have not found this to be an issue. > that in my rather cursory visit to the vendor's web site, > I saw no mention of ECC memory. I sure hope I just missed > that. ECC is implemented in the redundant storage area (as stated on our website). The redundant area is set aside because of both the ECC and the intrinsic NAND capacity required for bad block replacement and is the basis for the less than 4GB usable drive capacity. > Do we know if the device needs the OS to stick around after the > copy-to-flash command is issued? [A SLOG is best implemented without the added complexities of backup/restore.] To answer your question, the backup/restore feature is completely independent of the host machine and OS and can successfully complete without either. LEDs (red=backup/yellow=restore) are clearly visible on the X1 so that progress can be tracked, e.g. backup is initiated just prior to system shutdown with the intention of card removal at completion. Thanks, Christopher George Founder/CTO www.ddrdrive.com -- This message posted from opensolaris.org
