Well the VPSes are up and going which is a good thing, but the new HD is not seen.
We will now have to have another maintenance tomorrow at 9:30PM EDT. We suspect the power feed is bad on this hard drive as there is no way two have failed the new one being brand new.
I have edited the original topic to quickly reflect the new status, but keeping the old as listed for continuity. As before it will be about 15 minutes down.
ok with the change out of both the power and the data cable, the drive is working properly and being utilized now to rebuild the RAID array, we are keeping a watch on the rebuild at this time. WinVPS26 and 34 yesterday took about 2 hours to rebuild fully, we expect the same here.
IO may be a bit less during the rebuild, but not a significant impact.
The server is back up and raid rebuild resumed from where it was, however the error was caused by an error reading the source port, so we are watching it very closely.
it seems possible that there are so many bad sectors on the source disk, that it will not be possible to rebuild the RAID.
I am going to let it hang about 5 minutes and see if it can force through a copy of these bad sectors. Otherwise we will prep to move with backups, or remove the drive that is syncing to and do a live migration to a new node.
We have it up on one HD now, and will be migrating to WinVPS43 node for all clients on it. This one HD mode is not perfect and shown some issues in the last days, but it is better than hanging in the RAID rebuild so often.
The servers stayed up from the time we removed the 2nd hard drive with the rebuild making it hang, so we did not have to do any additional reboots and the only additional downtime during migrations were the few pings down for new hardware addresses to be seen by switching.