<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Nov 5, 2014 at 2:43 PM, Simon Hobson <span dir="ltr"><<a href="mailto:linux@thehobsons.co.uk" target="_blank">linux@thehobsons.co.uk</a>></span> wrote:<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
But back to the OPs situation, it's a bit "post horse departed, closing of stable door" suggestion - but RAIDing the OS directories would be a great protection against this problem. On the first error, I think MD would have thrown the faulty drive out of the array (for each partition/array it's involved with as an error occurs in that partition) leaving the good drive running.<br>
Recovery from that is a matter of partitioning the new drive, and telling MD to use it - then MD takes care of rebuilding the data onto it from the good drive(s).<br></blockquote><div><br></div><div>That's an interesting approach but I've never investigated software raid yet.</div><div><br></div><div>If I do it over drives of different sizes, will I lose the extra space on the bigger drive?</div><div>I have a bunch of drives in the machine and tend to upgrade them one by one, when they fill up or fail, with whatever the largest affordable drive is at the time of upgrade (nowadays 4 TB). So there will always be smaller and larger drives in there.</div><div><br></div></div></div></div>