You think that's bad.  A single hard drive has led to me spending 3 weeks rebuilding our entire network (I was at work from 30/12/2011 to 09/12/2011 before things were working well enough to have a day off).  Shouldn't be able to happen,  but it seems if you are using a single drive for the L2ARC cache on a ZFS file server and that drive goes a bit dodgy it starts corrupting things far and wide at a great rate of knots.
Surprisingly (or not if you know what ZFS is capable of),  we lost only 5 snapshots that were corrupted and had to be deleted out of a couple of hundred.  Everything important is still pristine.  I'm an even bigger fan of ZFS now.
				
			 
			
		 
			
				
			
			
				Zen wisdom: No matter what happens, somebody will find a way to take it too seriously. - obviously had KB in mind when he came up with that gem
Artificial intelligence is no match for natural stupidity
			
			
		 
	
Bookmarks