Hi Nick, Good report. We should be able to save your array without any trouble. On 08/07/2018 08:44 PM, Nick Gruszauskas wrote: > I have a RAID6 array that has been running nearly trouble-free for years. > This morning, however, the box of disks that contains all the drives for > the array was accidentally switched off while the host computer was still > up and running. I shutdown the host system, turned the JBOD back on, and > then booted the host system back up. Unfortunately, things did not return > to normal. It's ok. Abrupt cutoff means no updates to the superblocks made it to disk at the end. That's why they all show AAAAAAA (except the spares), with identical event counts (again except the spares). I'm gonna guess the spares were not in that chassis. > The array did not auto-assemble at boot, and I have been unable to get it > to auto-assemble once the system is up. I am also not able to manually > assemble the array, and force does not work either. I have worked my way > through the RAID Recovery article in the wiki and I am trying not to panic, > but I've reached the point where the next suggestion is to use --create to > recreate the array (and I'm starting to panic). I'm hoping someone can > offer some help. Do NOT use --create. Absolutely not! > When operating normally, my array has 7 disks and 2 spares. Every disk is a > 2TB drive. I am using mdadm 3.3.4 on CentOS 6.10. Yep. Perfectly visible in the partition exams. You should be able to reassemble by leaving the spares out. Probably don't event need --force, as the active devices are all consistent. There's likely a bit of filesystem corrupt due to in-flight writes not finishing, but should be minimal. fsck -n to check, and then fsck before mounting any filesystems. Don't forget to stop md127 before attempting the reassembly. Phil ps. Seriously... Do NOT use --create. -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html