Hi all! I did search the net and this forum, but couldn’t find any prior art matching my current issue, so please point me to what I might have missed, it this is already addressed somewhere.
My problem is that my Argon EON NAS (running two spinning-metal drives) has been unstable for a while now when rebooting and re-assembling the raid array I’m using the disks in. I’ve got a simple disk-mirroring setup, which as worked for the initial year or so since I set up the EON. But recently, one of the disks always gets assigned a new/different /dev/sd entry after reboot, and mdadm appears to be unable to reassemble the raid array. I can then manually fool around with mdadm to repair this, often resulting in the system spending a half-day with rebuilding the array, after next boot it’s likely broken again.
I tried setting up mdadm.conf to point to fixed /dev/disk/by-id/ entries, no luck though.
Any ideas what might be amiss here? I’m running Ubuntu server (always have, also when this used to work).