Re: RAIDframe: what if a disc fails during copyback

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

>> So you have drives A, B, and C.  A and B were live.  Let's say B is
>> the one that failed.  You reconstructed onto C and have been running
>> with A and C.
> Yes.

>> Now you have a new B [...].
> Yes.

>> So, you'd pull C, replace it with B
> No.  I don't pull C. I re-add B (I have lots of empty slots).

Well, I meant "pull" in the sense of "remove from the RAID".   Whether
or not that means "disconnect from the system" is semi-irrelevant.
But....

>> and initiate a reconstruct
> No, a copyback (raidctl -B).

This is an aspect of RAIDframe I don't know much about.  I'm guessing
here, but I would guess that this copies from C to B.  In that case,
for failures during the copyback....

If A fails, you have a copy on C, with no redundancy until the copy to
B finishes.  I don't know whether it is capable of realizing, after the
copy is done, that it could run on B and C; unlike most RAID levels,
for RAID 1 there's no reason in principle it couldn't - but I suspect
that RAIDframe isn't set up to do so.

If B fails, well, I assume you have to start the copyback over.  If
that's not the worst that happens, I would call it a major bug in
RAIDframe.

If C fails, you have to be running non-redundant, off A alone, because
the copy to B isn't finished.  In principle, it could finish copying
from A instead, but I suspect it's not capable of switching
mid-copyback.  My guess is you have to reconstruct from A onto B.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B