lionelhutz

I do do data a similar way, but I would not combine my media into a single share. I have 12 user shares but only 6 are exported so they appear on the network.

There is the option of setting the split level to manual. This actually means that once a directory is created it won't split. This means you'd have to create a series directory on multiple disks if you want it to split a season to a new disk. You'd have to create a new season folder on another disk if you want a season to split to that disk. This means that unRAID can place the directory on an initial disk automatically but gives you the ability to split the the directory if you so chose. Still, it's not hard to transfer series between disks if it's required. You can even use a plugin to do it. Use most free as the allocation method so new series will always start on the drive with the most free space to begin with. Doing this, I don't have to shift files around very often. I don't even think I do it once a year. I do keep a fair bit of free space on my server. Right now, my TV drives are all >700G free. Trying to keep series all together with mostly full disks can be tricky since a disk can end up completely full if you mostly fill it and then attempt to write a new season to it,.

I use 1 for TV so the whole TV series in on a disk. But, at most you could use 2 so that a season of episodes plus the season metadata all remains on a single disk. The series metadata could still end up wherever though.

Never pass through connected devices when you pass through a hub. unRAID tries to pass through the connected device but then "loses" access to it when the hub is passed through.

When testing, I have run 2 VM's with the same cores assigned and also run a Docker container with the same cores assigned as a VM had assigned. So, it doesn't seem like the CPU's are locked to the VM. You can test dynamically assign CPU's so that no cores are fixed to a VM. Basically, edit the XML and turn the section like this; <vcpu placement='static'>2</vcpu> <cputune> <vcpupin vcpu='0' cpuset='2'/> <vcpupin vcpu='1' cpuset='3'/> </cputune> into something like this. <vcpu placement='static'>2</vcpu> This makes for a VM with 2 cores assigned but it spreads the load out over all the hardware CPU cores available instead of the fixed cores you assign.

Here, 4T drives cost more per T than either 6T or 8T drives. If you have do them then they are the cheapest, but if you don't and are buying them then going larger is cheaper.

Yup, I posted /downloads --> /mnt/user/Downloads/ but it can be anything on either side as long the "/downloads" side matches in the different applications and the "/mnt/user/Downloads" side matches in the different applications. Of course, I'm only talking about the applications that talk to each other through an API like Radarr to SABnzbd. I deleted the incomplete mapping in LSIO. I certainly don't need it and I really have no idea why they would even put it in the template. Only reason I can think of is maybe they did it for a setup where SAB can does the post processing itself?????

^^^^^^^^^^^^^^ Don't use that because it will not work. It must be like below. BOTH the container and unRAID sides have to be exactly the same for it to work. FOR ALL /downloads --> /mnt/user/Downloads/ Point completed downloads in SAB to /downloads/complete and incomplete downloads to /downloads/incomplete. When incomplete downloads are in the Downloads share you can just delete this because it's pointless to have it in the container.

You have the containers mapped with /media pointing to your media share but the error says Sonarr tried to put the file into /tv. You either have to change the containers and replace /media with /tv and make sure the server side direction does point to the TV location on the server. Otherwise, set Sonarr to put the file into /media. In Sonarr, go to the Series page and then click the wrench icon for the Last Week Tonight series to see the path. If all series have the wrong path it'd be easier to just change the mapping. You should not need the media share mapped at all in NZBget.

You're a week too late. It was on sale cheap and had really good reviews. I'm not going to use one big HBA cards let alone multiples so the X4 slots don't matter. They will work fine for adding entry level video cards for VM's which is all I'd use them for. I do wish it had Intel Ethernet but the Killer chip is working just fine.

No, you're using it right. I'm just saying that stubbing the card won't affect it working or not. If you take out the stub line and uncheck the box in the VM config then the card would disappear from the VM setup window and you couldn't check it again. All the stubbing does is make it appear in the setup, not make it so you can use it. But,, you could add it back in by manually editing the XML for the VM without stubbing it and then it appears checked on the setup page.

I'm sure there was a reason but it wasn't obvious and after seeing the issue I'm just saying to be careful about using BTRFS with older hardware. Test it and make sure it's stable before jumping in with both feet. I'm not the only one that has had BRTFS corruption occur for no apparent reason. I'm using it fine now on newer hardware since I've done some upgrades in the last year.

You usually don't have to stub a USB card. It just won't appear in the other devices area to select until you add it to the XML manually. Then, if you un-check it and save the setup you can't just go back in and check it again but rather you'd have to add it to the xml manually again.

You'd think so, but the hardware was perfectly stable using XFS.

My guess would be a Sonarr setup issue. You have /media entered to be used somewhere, likely as the location for the completed files.

I had really bad luck trying BRTFS on my cache drive with a little older hardware like say 6-8 year old parts. The data would start to corrupt after a few weeks. I would recommend you test the stability by setting up a single disk and doing a few days of fairly intense reading and writing to it.

Can't read the title?

Just to make sure it's clear. A single BTRFS disk file system can detect data errors but it can't correct them. I've seen it confused that the BTRFS check summing will provide error correction, but it's only for the metadata on a single disk.

http://permissions-calculator.org/ The directories from files I wrote to the server via the user shares are drwxrwxrwx which is 0777 and the files are -rw-rw-rw- which is 0666.

It will fix new files. You have to fix the old stuff manually or using the new permissions script. I believe unRAID uses 0666 for files and 0777 for folders as the default.

Your plan sounds fine. I would also recommend you look at 8T drives. I know they cost more, but the $/T is about the best for a 6T and very close behind is the 8T drive. You will likely pay more per terabyte for a 4T drive then either the 6T or 8T unless you get them on sale for a deal. It would cost more to add 2 x 8T drives compared to 3 x 4T drives, but if you add another 8T drive the cost of that 8T will be less then adding yet 2 more 4T drives. RFS was a great file system. If it wasn't for the developers trouble it likely would still be one. But, it got pushed to the side and others have passed it.

Well no responses, but it seems to work fine.

Anyone successfully use this board? My server randomly hard locks up and I want to try another motherboard and this one is on sale right now.

Move step 3 to be step 5.5. I used rsync to do the heavy work instead of the plugin but either can work just fine. I did use the plugin to move a few things around later to free up space on some of my mostly full drives. I would also recommend you experiment with turning turbo write mode on or off. You get to it on the Settings tab under Disk Settings. It's called "Tunable (md_write_method):" and you set it to reconstruct write to use turbo mode. On my system it was way faster moving between disks with it set on. Don't worry about parity, it is always and immediately updated when you do any file system reading or writing.

Personally, I would put movies and TV in their own shares. I use split level 1 or only split the top level for all the media, movies, TV and music. That way, each Movie directory, each TV show directory and each artist is contained to a single drive. It's a personal preference to keep my media organized. I don't want the movie file on one drive and the metadata files on another drive. I don't want TV episodes for a single show spread over multiple drives. The issue this can cause is that continuing TV shows can attempt to over-fill a disk eventually. It's not a big deal to me. When it has happened I copy a bunch of continuing shows from a full drive to the most free drive I have and that's enough to keep it working for another long time.