[SOLVED ] - Help a newbie - 3 drives redballed

[Deryg]

check filesystem with repair? (no -n)

[trurl]

11 minutes ago, Deryg said:

check filesystem with repair? (no -n)

yes, you already did it with -n

[Deryg]

Moving files now. Next step is swapping out the cache drive then system will be good as new

[trurl]

5 hours ago, trurl said:

After that you can consider the array stable and it should be OK to enable Docker and VM Manager again.

You don't want to enable Docker and VM Manager until cache is ready

3 hours ago, Deryg said:

swapping out the cache drive

 

 

 

[Deryg]

shoot. 

 

They're running now. Also the damn parity check is running on schedule as I was moving files from the mounted unassigned devices drives. ETA for parity check... 12 days.

 

I assume that will speedup somewhat after the files move over.

 

i kind of figure Ill let all of that finish up, then I can disable those again as I put in an SSD drive to take over as cache drive?

At the moment, its running on the old Reiser FS spinner that it was originally. Will have to stop and shut down the array to put in the SSD anyways and I will have the room in the system to have them both the SSD and existing cache in as I will pull the old drives 5 and 6. Can't really stop the docker now as I am using Krusader to move the files

 

[Frank1940]

41 minutes ago, Deryg said:

They're running now. Also the damn parity check is running on schedule as I was moving files from the mounted unassigned devices drives.

IF it truly is a parity check, I would cancel it as running both at the same time will slow things down far more than running them consecutively. 

[Deryg]

All good. Files finished moving and the parity check now shows a sane 13 hours to finish. I had it scheduled to do a weekly parity check so It just kicked in

 

Kinda gun-shy to cancel anything after what I've been through

Jsut a matter of swapping out the spinner cache drive that's reiser-FS to SSD in appropriate file format once it's done and we will consider this unfortunate saga closed.

 

All the help has been truly appreciated, specially trurl with the timely replies and painstakingly walking me through everything.

 

Not sure how much data if any I will have lost between the old drives and the lost and found but it will be minimal

[trurl]

1 hour ago, Deryg said:

scheduled to do a weekly parity check

Why? Most do monthly or less frequently.

[Deryg]

I think that was the recommendation many moons ago, before dual parity. I set it then forgot it.

 

I recued the majority of the files from the lost&found although most were unneeeded. only a few thousand left to go through.

 

array seems rock solid now.

 

Heres a side question somewhat related.

 

Why do some user shares show as having some or all files unprotected. I invoked mover and it didn't change a thing. this understandably makes me nervous after this week's adventures in the land of data loss

 

[Frank1940]

Some of the shares (system, application data, domains) are cache only shares so they are not protected with your current cache configuration. 

 

I suspect that other shares have content written to the cache (where they are not protected) and later moved to the array where they will be protected protected.  (Unraid flags shares as in this condition of 'unprotected' even thought it is just one file!)

[trurl]

2 hours ago, Deryg said:

I think that was the recommendation many moons ago, before dual parity. I set it then forgot it.

I don't think more frequently than monthly has ever been recommended and I have been around since v4.7. And dual parity wouldn't change the frequency of parity checks. Parity is maintained realtime, a check is just to confirm that parity is correct. And parity sync errors are rare, or should be. If not, you need to get to the bottom of that problem.

 

2 hours ago, Deryg said:

some or all files unprotected

Any files not on the parity array or multdisk btrfs raid pool (cache) have no redundancy. Not necessarily bad, just something to be aware of.

 

You want the default shares

https://wiki.unraid.net/Manual/Shares#Default_Shares

on fast pool so Docker/VM performance isn't impacted by slower parity, and so array disks can spin down since these files are always open.

 

Even if you have redundancy for these it is a good idea to have backup of appdata. Since you have the plugin for that installed I assume you do.

 

 

[Deryg]

Thanks gents. I was aware of files not on the array. I was just curious as I had some shares that are array shares and I invoked the mover manually but it still shows. Not a biggy.

 

Also, thanks for the heads-up on the Parity. I'll be reducing the heat stress on my drives and moving that to monthly now. It may be something I'd picked up in the forums somewhere.

 

Now that my main issue is closed, I guess I need to figure out how to mark it solved.

 

I'll probably bug you guys to move my cache from ReiserFS spinner to an SSD that's a supported filesystem but that'll be in a different thread if I can't figure it out from resources available