Yesterday I started rebuilding a 4TB drive (2.85TB/4TB utilitized). The initial estimate was 141MB/s, and 20-some minutes in it was down to 120MB/s, which still would've completed this morning. When I checked on it at roughly 12:45am, 3-4 hours later, the speed had dropped to under 70KB/s. Rather than chasing it then, I just let it do its thing over night. The sub-12 hour projections have now turned into just over 1.5 years.
The "Cancel" is clearly labeled has leaving the array unprotected- . However, with it using the "same cable" SAS=>SATA breakout cable on the same SUPERMICRO AOC-SAS2LP-MV8, after unraid disabled the previous 4TB drive in that spot in the array (for a series of block-level write errors) it could've been a cabling item or a controller issue (went into service late Aug-2014). But, from this state during a rebuild... is there a way that would permit me to restart the rebuild from parity again?
Circumstance: I simply pulled the old drive, which is already 10 days "out of the array" (hadn't reconfigured notifications) so I'd definitely want to rebuild the drive rather than blessing it, and generating a new parity. The "replacement" drive had arrived yesterday morning, and with the array having been in an emulating state for over a week... I did not preclear it, restoring the protection for the rest of the array, prior to message with the "disabled" device outside of the system was my intention...
Based on the attached dmesg, the drive that 'encountered' the command error was one of the "old" 2TB green drives that has been in use since I was running 4.6. However, the other 3 drives on the SAS2LP-MV8 include the 4TB parity [which probably should be on a motherboard header...], 3TB data, 4TB being rebuilt.
unraid_6-1-9_syslog.txt
unraid_6-1-9_dmesg-grepping-ata1.txt