gundamguy

It's not 360GB because it's the default settings for cache pools is as a RAID 1 not RAID 0. The original point of adding BTRFS and cache pool capabilities was to add redundancy for cache drives so data sitting on the cache drive was protected from drive failure. You can change it to a RAID 0 pool if you want though.

<Sigh> I'm continually surprised at how much folks misunderstand disk sizes. Disk drive manufacturers have always expressed their disk sizes in decimal units ... e.g. 1000 bytes is considered a kilobyte; 1,000,000 bytes is considered a megabyte, etc. Computers speak "binary", where a KB is 2^10 (= 1024), a MB is 2^20 (1024 x 1024), etc. So a 120GB disk will have 120,000,000,000 bytes, but your computer will show this as [120,000,000,000]/[1024 x 1024 x 1024] = 111.xx GB It's not that your 120GB disk only has 111GB ... you're simply looking at two different ways of measuring the capacity. I think the confusion is entirely because Hard Drive Manufacturers are using a system of measuring that works in their favor but isn't actually how computers work...

The only windows machine I have in my home is a Windows 10 VM under unRAID I access from work with RDP. Am I safe assuming I have a strict control of the programs installed? or is possible to get infected with the port open for RDP? Thankyou Gus Just to be clear, this is a VM running on unRAID that you RDP into? If so you don't have to worry about the RDP part exactly... can you access your unRAID shares from with in your Windows VM (Meaning: When you RDP into Windows, can you navigate to your SMB shares?) If yes, then it is possible that an infection in that VM could spread, if no, then I don't think you have anything to worry about... yet.

http://lime-technology.com/forum/index.php?topic=35183.0 This should check for duplicates. Duplicates can happen, when you do disk to disk copy commands.

The answer is a definitive maybe... I'm not familiar with the USB to RJ45 adapter that you plan to use, but the first question I have is... is it compatible with Linux and are there drivers for it... even if there are I somehow doubt they are in the default build of unRAID so you might have to compile your own kernel with the needed driver. This is what I think the real issue is... if it works you should be able to use it... getting it to work might be a headache though. Then you would need to identify the port under your network config and direct OpenVPN at that port.

Not to mention, why? The server is going to be shut down anyway, so does it really matter if it's on the cache disk when it's shut down or on the array disk when it's shut down?

You cannot attach to a share with 'root' credentials and get 'root' access. This has been the case ever since v6 came out (cannot remember if it was also true with v5). If it is a public share you will be able to use the 'root' username, but the file will still be accessed with 'nobody' credentials. If it is not a public share then 'root' is not an acceptable username for connecting. Interesting, so this means the whole Cron Job to Read Only might work. You could also add a run now function and a "unlock" that gives it write permissions... giving you better control without having to mess with command lines.

Yes, it will only change when doing a manual balance. this really bothers me though... why does it say No Balance Found when there is a balance?

What are the advantages and disadvantages of each FS. Can you please explain the reasoning behind your recommendation. Thanks RiserFS is stable, but is end of life and isn't going to receive much future support (we had an issue about a year ago where an riserFS caused data loss because of it's lack of maintenance), the RFS file system check tool / repair is pretty great however. XFS is stable, widely supported and doesn't have the issues with nearly full drives like RFS. The XFS file system check tool is pretty good. BTRFS is the new kid on the block, it allows pooling (think Raid 1 at the file system level), some people have experienced corruption but not many... the BTRFS file system check and repair process is pretty new and people haven't had as much luck. It's improving constantly however. TL;DR XFS is great for data disks because it's stable and widely supported and doesn't have slowdown issues as drives get close to being full. BTRFS is used for cache to allow cache pooling, if it wasn't for the pooling feature I would assume people would use XFS for cache...

I am not sure if i would be using it for virtualization yet, thinking just storage, plex , a file share for family, but lately the kids want a minecraft server but think that would be better as a stand alone unit, With dual socket you could totally put that mincraft server in the same machine (as a VM).

Yes you should be able to do that in command line... BTRFS pools only work for the cache slots so I'm not sure it's using unraid is worth it at that point.

/mnt/cache/appdata is the better option since it bypasses the FUSE file system and just writes to the file system. Great, thanks for clarifying guys. How large do you typically make your docker.img? I'm not exactly sure what gets allocated to this space and I want to make sure there's ample room for future usage. For example, plex meta data wouldn't be stored in here correct? Since it'll be in mnt/cache/appdata/plex/... correct? Your question about Plex actually depends on how it's configured. But generally you are correct if your apps are "properly set up" that data will be outside of the docker image.

/mnt/cache/appdata is the better option since it bypasses the FUSE file system and just writes to the file system.

I'm not sure that's quite right, it doesn't let you do anything but raid 1 but I thought expansion and balance was handled by the unraid gui for Raid level 1.

If you can somehow mount the NAS drive over the network using unassigned devices then I think it could be accomplished with an rsync script. The real trick is somehow mounting that NAS drive.

doesn't unraid handle this on its own?? It depends, if you reference /mnt/user/ on a cache enabled share... yes I think so. If you reference /mnt/cache/ then I don't think it does, because it's not part of the fuse file system.

I think this is a good question. Part of the issue is that you can balance in a lot of different ways, Raid 1 2 disk, Raid 1 4 disk, Raid 0 2 disk, Raid 0 3 disk, etc, etc, etc. Now officially I think Limetech only supports Raid 1 with 2-X Disks, but I think the major problem is how do they handle all the other options that can exist. I think at minimum it should prompt you on if you want to balance whenever it notices the cache pool has changed.

My guess is you have your split levels, or min free space settings set in such a way that shfs attempts to put it on a disk that doesn't have space.

Always the best policy. Let us know how it goes!

IDK, I'm using BTRFS now and have been for several months (not a full year) with no problems. Also I saw that BTRFS has more guidance on using btrfsck and recovering from errors now... (as in only do btrfsck as the last possible thing, and use btrfs restore first as it's non-destructive) so maybe recovery tools are progressing.

What worked for me was disabling SMB 2&3, forcing the connection to SMB v1. Guide from Microsoft. The relevant parts. And if you wish to reenable SMBv2 and SMBv3 at some point in the future when compatibility issues are worked out you do this: Notes You must run these commands at an elevated command prompt. You must restart the computer after you make these changes. This might not do it, but it's worked for me and a few other people.

Based on the directions on Plex's website this can be done... Im pretty sure the plex plugin from Plex's website uses "Plex" as the user in linux, I don't know that linuxserver.io does. I'm not an expert on docker so I don't really know. The hubpage for the docker seems to imply that it's running as "plex" instead of "Plex"

Are you using the same plugin or docker you were using on machine one for Plex on machine 2? Did you pass it the exact same configuration? My understanding is that if the /Plex Media Server/ is moved properly Plex will have no problems. So two things could be messing this up. 1) The directory tree isn't correct and it's not finding /Plex Media Server/ because the configuration is slightly off. 2) The permissions aren't the same and you aren't running it under the same user (making it so even if the files are in the right place Plex can't read them) so you might need to change the permissions to be readable by Plex. Edit: Plex has a guide to this as well. Depending on how this was installed your mileage my vary on what exactly you have to do.

I would consider looking at the VPN plugin, I think it has good instructions.

Interesting way to put it... he isn't being hacked, it's just that people (actually automated scripts) are trying to attack his system but... You are correct that you can make it harder for them to use automated systems to break into your server, but here is the most important point.... unRAID doesn't disable password login by default and doesn't use private public key log in by default... so the reason people say... don't do this is because we don't like to assume that they have done this... Not to mention that there are smarter safer ways to go about this.