JonathanM

If you typed that accurately, you will quickly fill up RAM. Can you post a screenshot of the docker settings for this container please?

The link you posted is actually more up to date. The unraid team is rather small, they spend most their time on the software and not the docs.🙄

Which is why so many of us use Unraid. It's that good.

https://wiki.unraid.net/Parity

For the main array, Unraid with single or dual parity is the only option. For the cache pool, any valid BTRFS RAID mode can be used. Unraid is not RAID, each disk has a valid partition all to itself, nothing is striped. The parity information is only on the parity drive(s), not spread across the drives.

Cool. As long as you have the data physically elsewhere. Drives dying are only one small percentage of data loss causes, especially with unraid running good hardware. I see MANY more instances of direct user or configuration errors causing data loss.

After a drive is disabled (red ball) it must be rebuilt. Nothing to do with acknowledging an error. Stop the array, unassign the offending drive from the parity slot, start the array, stop the array, assign it back to the slot, start the array and let it rebuild.

No restrictions on assignments. You could have 1 data drive and 11 cache drives in a pool, 12 data drives with no parity, 1 data drive and 11 unassigned that you could mount and work with manually or with the Unassigned Devices plugin. The only limit is no more than 12 attached storage devices when you start the array. It's been some time, I wouldn't call it recent. However, in the context of unraid version releases and the definition of soon™, I suppose it's recent. 🙂

Yes, as long as the controller connected to that slot supports hot swap.

That sentence doesn't parse for me. Can you explain more, and post screenshots of what you mean?

You certainly can rebuild both at once, however, the safest option would be one at a time, with a correcting parity check showing zero errors after each replacement. Totally up to you how much risk you want to endure.

Perfectly correct, however some motherboards won't boot without a GPU, or need a BIOS option changed to boot without one. It's always best to have a GPU available to temporarily use to troubleshoot.

This is the more interesting part of your post to me. How are these drives connected? SAS would be ideal, eSATA acceptable. The answer to your direct question is in the pricing link at the top left of this page. https://unraid.net/pricing

Yep. https://unraid.net/pricing

1. Yes, hardware permitting. 2. Yes to the VM's, the NAS function utilizes entire drives, not just partitions. 3. Very easy to very complex, depending on which OS and what kind of hardware allocations, virtual or real device. The virtual disk can be just a file, so no partitioning needed. 4. Yes, manually. They can be scheduled with scripts. Assuming the USB devices have unique ID's you can attach them like you say. 5. Yes, as long as you have a spare drive or two that you don't mind erasing for use with the trial. You must not assign any drives to the array that you don't want erased. Unraid loads into RAM from the USB Boot stick, it doesn't use a hard drive for the OS. If after you trial Unraid and decide you like it, if you wish to use the entire drive that windows was originally installed on as an array disk, that is easily doable. 6. No. That being said, Unraid uses standard linux KVM as the VM engine, so any guides for migrating foreign VM's to KVM can be adapted to Unraid.

<snrk>🤣

Did you download the diagnostics zip file BEFORE you rebooted? If not, then if it happens again, be sure to collect diagnostics before you reboot.

Typically that is caused by physical connection issues. Check the ports for bent pins, try different cables, etc.

Depends on the specific torrent and tracker. Very popular torrents on some public trackers can work ok, because your client can connect to other clients that are actively transmitting data in both directions. If there is nobody you are currently downloading from that needs data you have, you will not be able to seed at all. Responsible torrenting involves allowing others to download from you as well as just downloading from them, and all private trackers that I am aware of will not allow you to participate without seeding and keeping a positive upload ratio. It's definitely not a myth, it's a fact about how the torrent protocol works. Many torrent clients actively seek to prioritize traffic to clients that have healthy connections, so if you aren't connectable, you will have a hard time downloading on many lightly populated swarms.

File system corruption doesn't equal drive failure, and replacing a drive isn't going to fix the file system, regardless of whether or not the drive is bad.

Yes. It will ping you immediately as long as you have it configured properly and the server has internet. I would suggest you research this further. It's not just 2 drives that create the protection, it's ALL the drives, data drives included. When a drive fails ALL the other drives are needed to emulate and reconstruct the failed drive. Adding the 2nd parity allows 2 different drives to be emulated/recovered, but the rest of the drives must all be perfectly healthy to reconstruct the failed drives. Also, it's not file level parity, it's device level parity. There is no way to reconstruct individual files, only entire drives, filesystem included. A corrupt filesystem or deleted files cannot be addressed by parity protection. This is why I stressed current backups are needed on an ongoing basis. Why not use these drives to build out your unraid? That way you can keep the windows server up and running during the whole process?

I believe those files, DS_*, are the result of accessing the share with a MacOS, and can safely be deleted.

Not for general functionality. However, after I ran it for almost a year without it, I decided I really owed them something, and purchased a "lifetime" license. I was a little bitter when I read the fine print and found out it's not necessarily a lifetime license, but in reality it's similar to the Unraid perpetual license, where the authors could decide at a later date to start charging for upgrades, but several years later and Emby hasn't asked for anything additional. I've definitely gotten my money's worth with all the use I and my friends and family have gotten from it. I recommend reading the binhex emby support thread thoroughly in its entirety if you are truly interested in premier. Paying a developer for great software encourages more development, so I definitely recommend purchasing premier if the basic functionality appeals to you.

Nextcloud is the ongoing project of the original owncloud developers IIRC. Similar but now different in execution. I currently use it exactly as you describe and it works great. The key is the external storages plugin.

No problems with those drives, HOWEVER... They cannot be assigned to Unraid and keep the current data on them. Any disks assigned to Unraid will be erased and set up with a linux compatible file system, XFS or BTRFS. You will need a way to migrate the data, either from your backups, or if you don't keep a backup, you will need to purchase additional disks. Drive replacement redundancy is not a substitute for backup anyway. Data loss is much more likely from human error or malice than equipment failure.