Replace original removed HD after failed disk upgrade?

Do you have a backup? If not, you should make one before you go any further.

If you removed disk 1, you could power down the NAS, put disk 1 back into the NAS, and remove disk 2. Then power up the NAS. It should boot to a non-redundant state (and holding the information when the drive was removed). Hot-inserting disk 2 would then resync the duo.

If you removed disk 2 it might be more difficult (it shouldn't be, but there have been several postings from folks who've had trouble booting up with only disk 2 installed). Though the protection mode is raid-1, the disks aren't precisely the same. Disk-2 has no partition info on the data partition for one thing.

Since expansion didn't complete, you could possible hot-insert disk 1 (trusting it to resync to disk 2). I wouldn't do that w/o having a separate backup.

Many thanks, Stephen. Wise advice re the backup! I'd started a backup across the network earlier this morning to a shared USB drive as a precaution. Looks like I have about 24 hours to consider my next move while the data copies across…

It was disk1 that I removed. If I follow the procedure you outlined in your second paragraph ("If you removed disk1, etc…") would the resulting volume (once resynced) match the old state of disk1 (i.e. as it was when it was removed) or the newer state of disk2?

if you install disk-1 powered down (slot-2 empty) you will have the old state. Don't leave disk-2 in, as then the NAS will think it has a synced RAID array when it doesn't. hot-inserting disk 2 will sync disk 2 to the disk-1 (keeping you at the old state).

Hot-installing disk-1 might possibly fail (since the new disk-1 is 2 TB), I am not sure. The volume size is still 500 GB, so the xraid scripts might be ok. If it works, that would sync disk-1 to disk-2 (so you'd be in the newer state).

The basic principle is that every hot-insert is treated as a new disk - the NAS runs a disk test, and if it passes reformats it, then resyncs the raid array. But if the NAS is powered down during the insertion, then if it recognizes the format on boot-up it will treat it as an existing disk in the array. If it doesn't like the format, then it will usually treat it as a new disk (though it might simply fail to boot).

Thanks again. Your explanation is crystal clear!

Hi again, Stephen. The Duo has been backing up to a USB HD connected to the front port since 9:40 am on Wednesday (i.e. +/- 50 hours ago). Is this a normal length of time for a first backup of around 430 GB, would you say?

That sounds slow (~2 MB/s).

Did you enable "Enable fast USB disk writes"? Also, are you able to see how far the backup has gotten?

A ha. No, I hadn't. I wasn't aware of that option. I imagine I should play it safe and not tick the box mid-backup. Will enable it as soon as the backup is complete.

I had been looking for some sort of progress indicator in the Backup section of FrontView, but of course there isn't one. However, I've mounted the USB drive on my Desktop (something I didn’t think would be possible mid-backup for some reason). It's currently up to about 350 GB, which, by checking the Info at intervals, I see is increasing at a modest pace. So, it would seem logical to infer that the backup is (a) slow, but (b) still making uninterrupted progress.

(And - thanks again!)

If you do enable it, be very careful to unmount the USB drive before you remove it. If the backup is scheduled, it is good to protect both the drive and the NAS with a UPS.

This painful saga rolls on…

Both Seagate 2TB ST2000DM001 drives failed after or during initial sync. They’re going straight back to Amazon, and I've ordered a pair of Western Digital Red HDs. For reference for anyone searching this forum in the future for compatibility experience, both Seagate drives were manufactured in China, not Thailand. Firmware on both was CC26. Based on my experience, I wouldn't go near that model for use in a Duo v1 if I were you.

And now we get to the thorny bit…

I hot-swapped the 2nd "dead' Seagate out of SATA channel 1 this morning and replaced it with the old 500GB drive from that channel. The smaller HD initialized and then the Share disconnected. RAIDar could not locate the Duo, although the Duo's power light was steady, and the green LED was flashing for the new HD and steady for channel 2 in which has remained the second of the 2 original 500GB drives all along. After half an hour of refreshing RAIDar to no avail, I pulled the power to the Duo.

On reboot, Frontview reports the 500GB drive as "UNKNOWN [1862 GB]". Clearly something is amiss here.

So, I shut the Duo down from Frontview. Then I removed the mis-reported HD from slot 1 and powered the Duo back up with no disk or caddy in slot 1 at all. And still Frontview tells me that there are 2 disks in the Duo, and that the one in slot 1 is "UNKNOWN [1862 GB]".

I've made another USB backup in case everything goes south before the WD Reds arrive, but does anyone have a suggestion for how to persuade the Duo that it does indeed only have 1 disk in it, as I'm concerned that this might mess up installation of the Reds when the time comes.

(Experienced users, if you think this merits a new topic I will happily make one, of course.)

The dead/removed disk should be shown but with a yellow LED till it is replaced with a working disk of at least the same capacity (in the case of the Duo v1 you can't install a larger drive as 2TB is the max supported).