gundamguy

1) Yes looks good. It's not really like RAID5 for a lot of reasons, but you are set to survive a single disk failure. 2) This id going to depend entirely on the logic you set up for your user shares, like which allocation method (High-water, Fill-up, etc.), Excludes/Includes, and Split-levels. So it doesn't have to do that, but it can if you want. 3) This is a good question. Hope someone else can give more insight into this. 4) A user share is a file system in user space that aggregates multiple disks into a single share. A disk share shares only that one disk. /mnt/user/movies is a user share and /mnt/disk1/movies is a disk share. The user share would read all of the /mnt/disk1/movies for each disk and aggregate it. (Caution, mixing user shares and disk shares when doing move or copy operations can result in data loss) Most users use user shares exclusively. 7) Adding an SSD cache gives you to things, the illusion of quicker writes to the array since writes to cache enabled shares would written to the SSD and be moved to the array at a later time. The second thing is the SSD gives you a location to put VM/Dockers/Application files (Plex Metadata) that won't spin up your array disks.

If you are on v6 post your diagnostics file and we can look to see if there is anything noticeably wrong.

Is it possible that there was a mix between /mnt/user/Movies and /mnt/diskx/Movies in one of the automated programs that could be clobbering the data? IE: Is there any possibility that the corruption is due to the user share copy bug?

Is there a reason you want / need this?

Your internet speed isn't really relevant because this should be all done across your LAN. 10mb/s is what I would expect of a network connection working at 10/100 not gigabit speeds. Your unRAID network adapter is gigabit it seems, so check to make sure you are using cat5e or better cables, that any switches or routers in between are also capable of gigabit speeds. I think you have a weak link somewhere that's slowing it down to 10/100.

I think you are correct, and because your array hasn't been started you don't get the controls for spin up or spin down. (I didn't know it worked that way...) The main takeaway is that this isn't unexpected behavior or a problem really. unRAID is just trying to conserve power by spinning down disks that haven been inactive for a set amount of time. unRAID will spin up the drives automatically when needed and there is a setting you can change to determine the spin down delay, or even set disks to never spin down. When disks are spun down they show up as inactive (usually, in your case they might be unassigned) and you lose the smart / temperature status, since requesting that information would spin up the disk.

Not sure about the unassigned drives actually, but on the main tab the disk should have an up or down pointing arrow next to the status indicator. Clicking that arrow will either spin the disk up or spin the disk down. (This might be true of unassigned disks as well I don't have any right now so I can't look to verify)

I support this if a proper warning is given about redirecting it to the flash.

I don't know guys, I'm starting to feel it coming on.... JonP has been running around the forms more lately and squashing a few bugs...

Based on what your trying to do, I don't think you'll have any problems, but the reason disks exports are by default off is to reduce the risk of the user share copy bug. Just keep in mind that if you copy from a disk to a user share (or vice versa) that contains that disk you can have data loss when a collision happens. I highly recommend turning off disk exports when you are done with your move.

That just means the disk is spun down to save power. When it spins down the status goes grey and the smart report drops off line (because it's spun down.)

Lime Tech. The main point is that we don't have the beta release of 6.2 yet... but maybe soon?

I could have sworn they were exact, but after a closer look, you are right. One of words had a lowercase i. Thanks. So know that that's been identified, how do I get both folders to show back up so I can combine the two folders? Any idea? If you rename the folder with the lowercase i to have an I instead, you should stop having this issue.

can't you also upload the .opvn file to your IPad with your inline certificate, and settings via ITunes? It's a mult-step process, but basically you configure the OpenVPN server, then create a Key (or keys) for the client(s), then open the proper ports so it can communicate with the internet. Then give that Key and server config info to the Client application so it knows where to connect, and what other settings are required. The actual process to do this varies based on the clients you are using. I'm sure you aren't the first to ask questions though so read though Peters thread, I'm sure it'll help you figure it out.

I don't know that a full history is stored anywhere, but the before and after smart reports should be stored on the flash drive. I don't recall the exact folder but there should be one with the before / after smart reports at least.

Yes but be careful. Make sure you move from Disk1 to Disk2. Make sure you DO NOT move from Share to Disk2. If you mix and match user shares and disks shares in your move operations you can clobber the data and lose it.

My understanding is that by default mover invokes rsync with the -t flag so it should preserve modification times.

What Squid said. Additionally if you are curious User0 exists to make Mover work. Mover rsyncs from User to User0... You can take advantage of User0 to do a user share write that bypasses the cache and writes directly to the array, if for some reason you wanted to do so. Actually mover rsyncs from cache to User0. That's correct. That would be a massive rsync otherwise. Seems like I should get some sleep, since I am making simple mistakes.

Your best bet is attempt what bobekdj said and try mounting it with unassigned devices. But depending on what happened to the disk it might not be mountable. Edit: There is a lot of unknowns here, like what the original file system is (assumed NTFS since it was being used with Windows), if that's the issue, if it's a hardware issue... etc. You've not given us much to go on, which is why it's hard to tell you what you can do.

What Squid said. Additionally if you are curious User0 exists to make Mover work. Mover rsyncs from Cache to User0... You can take advantage of User0 to do a user share write that bypasses the cache and writes directly to the array, if for some reason you wanted to do so. Edit: Corrected.

Why is the plugin a third level choice. I have it working well. Only issue is connecting from China to my unraid box in the USA, is no longer sufficient to get access to Facebook, YouTube, Gmail etc. But I can still get access to the server and the rest of my LAN. Openvpn used to allow my Gmail to work in China. The only reason I consider it a third level choice is because it's running on your unRAID box so if you need to restart or you have a power failure that causes a restart your array will go down and your connection, making it a little worse for doing some tasks with unRAID. OpenVPN as a plugin works fine so it's not terrible if that's what you have to do, it's just a matter of knowing the limitation.

That's good news. At least once set, it won't randomly switch back unless you mess with it. Did you determine this through your own experience, or has Limetech (or someone else) posted somewhere else that it is persistent? From what Tom posted in the above link, he made it sound like it was not... After reading that 2-3 times I'm pretty sure that what Tom is talking about is the code that shows up on the cache page (that you can modify and run) and that is run when a drive is added to an existing pool (this seems like the bigger issue now...)... Are you having Plex make thumbnails, or do you just have that much media?

The last I heard, about middle of last year, emby (and Plex) for that matter both use ffmpeg to do there transcoding, and aren't capable of taking advantage of GPU resources. So even if you added it emby would not benefit from it. Your best option is to upgrade your AMD APU. Rule of Thumb is you want a CPU with a passmark score of about 2000 for each concurrent HD stream. (So for 2 streams about 4000). If you are doing anything else processor intensive will cut into that as well, so having some overhead is a good idea too.

you should check the S.M.A.R.T report to see what's up. Based on the fact that of your 450000 errors it appears there were only 13 good writes... I want to believe that there is some sort of issue with the cable.

Print this screen shot out, and just make sure you match up the serial numbers to the same array slot once you reboot and you should be good to go. (Which drive is connected to which SATA port doesn't matter, just the assignments)