resinch a dieing disk or add new disk & resinch?

I probably wouldn't run the destructive test on HD4, since there is no reason to think it has failed.

If HD1 was partially synced, then I would certainly put HD4 into slot 1 and re-establish redundancy. HD1 is still perhaps questionable, and you don't want that resync to fail.

Then I'd put HD1 into slot 4.

On option 2, the NAS would fully sync one disk at a time. I see no point to steps (c) and (d). Adding one disk at a time and waiting for completion is safer, as you are always clearly aware of what the NAS is trying to do.

BTW, there is no benefit in trying to order disks by age, and over time it is not sustainable anyway.

Just as we thought we were on the finishing straight....
Look at these printscreens.
https://www.dropbox.com/s/1jnr5w92vphn8r8/Victoria%2C%20from%20cd14%28Gaelle%29%2CHD1%20256%20just%20the%20ABRT%20bit%20Finish%20OK.jpg
https://www.dropbox.com/s/3sl7xg4k399nxzs/Victoria%2C%20from%20cd14%28Gaelle%29%2CHD4%20256%20just%20the%20ABRT%20bit%20Finish%20OK.jpg
Two different disks, two near-identical results.You can read the HDs IDs on the black bar on top. W1F0WXJW is HD1. Z1F3362S is HD4.
HD4 - brand new, mint out of the bag 2 weeks ago ready to go into NAS, never spun until yesterday when I started testing.... fails Victoria EXACTLY like HD1. I tested it "for a laugh"... and I am NOT laughing.

Now I just started running Victoria on the entire HD surface, it will take its time (and Victoria only test 49% of the disk each time, don't ask me why).
These results also questions Victoria.
At the same time, HD1, passed by SeaTools and Lifeguard, was logged as a "fail" by NAS, so I really do not know what to think.
I wish I could say "I give up"

I cannot preview the images, but I can open them in a new tab.

I have no experience with Victoria, I agree that perhaps it might be generating some false results.

Have you tried booting the NAS without disk 1? Do that, and see if you see your data.

Sorry, this morning I restarted some HD4 scanning before reading your mail, and it hasn't finished yet (69.6% done).
Sure, I can start NAS without HD1, but I don't know "how" I would know if my files are accessible: I have (say) 4Tb of data, which is everything from Movies, TV cooking programs, all my digital photos (and video clips) since ~2000, work stuff, backup data for various PCs, laptops, netbooks, phones etcetera.
I don't know if seeing a folder structure means that the data is accessible, or I may see a folder, but when I click on it, the folder will open, but I can't open the file. My experience of file damage, is that a file which has got corrupted "is there" (sometimes taking the exact amount of space) but does not open, although some types of files do open partially (like some pictures, or movies which don't run the full length).
How would I know if I had data loss? Apart from opening one zillion files, I would never know.
Also you said that NAS would do a "volume scan", and that "some data loss may occurr" - which is what I would wish to avoid.
So we got into HD testing, and, apart from Victoria which we speculate that could be generating "false positives", HD1 has been perfect, all test passed, SMART showing no evidence of ever a problem, etc.
Basically I thought that we would be satisfied with that, so I cold-refit HD1 and let it finish the synching which it started doing. I stress that the resinch did not stop or fail - I interrupted it as soon as I realised that the NAS, which was rebooted from the netbook in another room, was back online but on a "waiting (or so it appeared) state". I rushed to the NAS, I saw that it was re-synching, and I immediately switched it off (soft option).

On the other hand,

If HD1 was partially synced, then I would certainly put HD4 into slot 1 and re-establish redundancy. HD1 is still perhaps questionable, and you don't want that resync to fail. Then I'd put HD1 into slot 4.

so obviously the initial "fail" flagged by NAS is very much on your mind

And you obviously mean that you would run Seatools destructive test on HD1 before refitting in bay4, I presume.

The RAID array is still functioning if you can start the NAS w/o HD1, and you can see folders, and access some files. There of course could be some undetected corruption, or even hardware issues with the remaining disks. If there are, then there is nothing we can do to fix that anyway.

On the other hand, if it won't boot, or if you can't see any data then I'd say it is unlikely that you can get any data back.

And yes, I am still concerned about the initial "fail". Personally I would run the destructive seatools test (or its equivalent) on HD1. The test simply writes zeros to every sector. After it runs, I look at the reallocated sector count. If you don't have a freeware tool to look at that, you might download acronis disk monitor.

OK, totally confused.
I switched NAS off few days ago. Now I just clicked on its icon on network resources .. and went in the NAS room to power it up...... to find it ON, door open as I left it, saying "Disk 1 fail, C: unprotected" !!!.
It booted up by itself.... I did not even know that it would answer network calls if it was off (I ticked a box for that, but that was because I want to have access while away. I always assumed that it needed to be "ON", but maybe asleep). Why do I ever bother to leave it on, if it is going to power-up like that?

All files seem to be there. But as I said, if I remember right and I had 38% free on 3x3Tb HDs, that is 62% of 5.6TB = 3.5TB of data, I opened a couple of directories and clicked on few files, apart from the delay because I am on the WiFi, everything came up ok, but it is near-impossible to know if I had data loss until I hit a corrupt file.
OK so next move is hot-insert HD4 in bay1.
Then run Seatools 2.23 for dos (you suggested the vendor's software) and write 0s (now I know which of my hiren disks has a working copy of SeaTools). I also have Acronis Disk Manager, but for partition work I used DiskGenius (now renamed Partition Guru or something like that), which also has a "verify and repair bad tracks" tool.. silly me, I should have tried that instead of Victoria...
But then, it might have said that HD1 is OK, just as everything else has done so far - apart from NAS.... and the "temperamental" Victoria

Quick update. As expected HD1 passed the Surface Test from Disk Genius and EaseUs (which took 6h 50minutes).
I knew that I had various version of Acronis Disk Commander, but none of them seemed to work, so I downloaded and then installed a newer version but that also had no diagnostic tools on it.
Few minutes ago I realised that you said Acronis Drive Monitor which I have just downloaded.
Last night I eventually hot-inserted HD4 in bay 1, NAS checked it briefely and after a minute or so started resynching. It took 12h40m and now it is 100% synched. Volume is 5.9TB, 1.8 TB availlable.
I have not (as yet) encountered any evidence of data loss, but I have hardly accessed NAS, apart from checking some video files. These files were part of a set (50Gb of files) which were probably the latest files copied to NAS - infact they might have been copied at the same time that the fault occurred.
I will run the Acronis diagnostic test on HD1 and will report.
Thanks again

All Acronis Disk Monitor does is report the SMART stats - which is handy, but not a diagnostic.

yep, I found out that.... back to SeaTools for the HD1 "write zeros" destructive testing. I'll report when done