On Thu, Apr 6, 2017 at 7:31 PM, John Petrini <jpetr...@coredial.com> wrote: > Hi Chris, > > I've followed your advice and converted the system chunk to raid10. I > hadn't noticed it was raid0 and it's scary to think that I've been > running this array for three months like that. Thank you for saving me > a lot of pain down the road!
For what it's worth, it is imperative to keep frequent backups with Btrfs raid10, it is in some ways more like raid0+1. It can only tolerate the loss of a single device. It will continue to function with 2+ devices in a very deceptive degraded state, until it inevitably hits dual missing chunks of metadata or data, and then it will faceplant. And then you'll be looking at a scrape operation. It's not like raid10 where you can lose one of each mirrored pair. Btrfs raid10 mirrors chunks, not drives. So your metadata and data are all distributed across all of the drives, and that in effect means you can only lose 1 drive. If you lose a 2nd drive, some amount of metadata and data will have been lost. -- Chris Murphy -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
