tshorts

If the main use of a share is to download files with torrent-protocol, would you use cache disk? Why? Why not?

It's not the traffic load i want to balance (which is how I always used the expression 'load balancing', it's the one external listen port -> several internal server ports. Kemp call it load balancing which I think is a bad name for it, but I don't know any other word for it). So replacing the hardware router with a virtual one, or implement vlans wouldn't solve the 'problem'. Perhaps reverse proxy and load balancing should be my search terms. Found Nginx, might solve the problem. I have to look deeper into that.

Are there any load balancers for unraid? Like Kemp, but free, and as a docker, plugin, or command. Load Balancing is quite a broad expression in network, but I think that's what it called. So narrow the explanation: All request to https://mydomain.com goes to the load balancer. The load balancer checks the full address and directs it to the different internal servers based on the address. nextcloud.mydomain.com goes to the nextcloud server. ftp.mydomain.com goes to the ftp server. plex.mydomain.com goes to the plex server. etc. Instead of my setup now where redirects is port-based and done in the router: mydomain.com:532 goes to nextcloud, mydomain.com:21 goes to ftp, mydomain:8324 goes to plex. (Just examples not my real ports, servers, or domains)

Removed them from console now, config is now clear. Big thanks!

Set up nextcloud when I just installed unraid to test but never used it. Want to use it for real now, an forgot all testpasswords so I removed mariadb and nextcloud docker to set it up from scratch. But it's still the old settings left when I install it. E.g. root password for mariadb, and nextcloud thinks I've already created an admin account. How do I clean this?

That did the trick. Big thanks!

Just upgraded from 6.8.x to 6.9.2 - and noVNC started to give following error. noVNC encountered an error: Uncaught SyntaxError: The requested module '../core/util/browser.js' does not provide an export named 'hasScrollbarGutter' http://192.168.10.50/plugins/dynamix.vm.manager/novnc/app/ui.js?ts=20200718:11:35 SyntaxError: The requested module '../core/util/browser.js' does not provide an export named 'hasScrollbarGutter' Is it a bug or do I have to do some tweaks after upgrade?

I'd like to disable my dockers (right now only Transmission), and then Spind Down the drives at lets say 00:00, and then start the docker again at 06:00. Right now I do it manually, but I forget it 3 of 4 times so I need to start my desktop PC up again and pause the docker, and stop the spinning. Reason is my new harddrives in the unraid server are to damn loud. I can't find any scheduling built in to unraid, but I guess it's possible to script? Or maybe even an app for a n00b like me?

+1

@ChatNoir Version: 6.8.3 I'll try take an extra look that everything is shut down. Nothing that tries to access it if all plugins, dockers, and VM:s is shut down? Or do I need to think of something else?

After I installed new hard drives my server became quite loud. The new hard drives are much louder than my old ones. Can I somehow pause the drives from spinning for like 8h? With built-in or addon. I've tried - "Spin down", they stop spinning for a couple of minutes and then start again. - "Stop array", they start spinning after stop is completed. - "Start array in maintanance mode" + "spin down", they stop spinning for a couple of minutes and then start again. I want the hard drives to spin down, and stop them from spinning up again for a few hours, without shutting down the power for the server.

Thanks! Will do.

Thank you, that worked. Disk 3 wasn't even assigned after that. Was that because it was set to "no disk" when I did the "new config"? I wonder as I will replace 2 smaller data disks with 2 larger data disks and one larger parity disk, so I will have to do this again the upcoming days after some zeroing and copying.

I have 2 data disks in the array, 0 parity disks. Added a third data disk by choosing the hard drive in Disk 3. Got "unsupported partition layaot" on Disk 3 when I started the array. Started preclear it, and remembered that's the disk I want to use for parity. I then chose "no disk" on Disk 3 to remove it from the array. Which gave me error "Missing" on Disk 3, and error "invalid configuration - Too many wrong and/or missing disks!" by the greyed out start array button. I found something about this when googling, https://blog.linuxserver.io/2013/11/18/removing-a-drive-from-unraid-the-easy-way/ but it's from 2013 and felt so complicated and risky for something that hasn't even been active in the array. This method will not delete any files even without parity?

I try to remove a disk from the array to use it as a parity disk instead. But when I change the slot to "no disk" the disk slot says "Missing" and the array can't be started due to "Invalid configuration". The disk was new and empty and connected to the array for about a minute but didn't start with it due to "unsupported partition layaot". It's now being zeroed while the array is down.

Was thinking. Perhaps it's just HDD spinup. Possible to have the indexing on the cashe drive? Or is it already there?

It's same workgroup, and I've tried entering the server many hours after pc-bootup and it's the same. And a few minutes after the first try it works (until shutdown or several hours, not sure). The network is there, and the icon for the server is there. But the resources unavailable.

Thanks. It's set to "Yes", and dynamix show it's local master (with server uptime at 20 days). I added the commands from your post and rebooted the server and it started up as local master. Clicked the server from my desktop and it was still fast (I "initiated" the server when I took the screenshot above and haven't rebooted my PC since). I rebooted my PC - and I got the error message again. It seems the issue is on the Windows-PC end, and not on the unraid end. Like SMB is idle until initiated or something like that.

This error always come up when I try to access the server through Network -> Server in explorer for the first time after my desktop PC has been rebooted. After that I have to wait a minute or two before I can access it. Can I somehow make windows initiate connection to the server at startup, so it's initiated and works immediately when I want to use it? I don't have the issue if I try to e.g. run a movie-file directly from transmission, it's only browsing that takes time.

I'm trying my first passthrough for a "guest gaming PC" on my unraid server. The Windows VM boots when I only have the NVIDIA Graphics card, and NVIDIA Sound card enabled. But when I try to passthrough a secondary soundcard, the one integrated on the motherboard, which is the one that will be used, the boot fails. IOMMU group 10: [8086:a2c8] 00:1f.0 ISA bridge: Intel Corporation 200 Series PCH LPC Controller (B250) [8086:a2a1] 00:1f.2 Memory controller: Intel Corporation 200 Series/Z370 Chipset Family Power Management Controller [8086:a2f0] 00:1f.3 Audio device: Intel Corporation 200 Series PCH HD Audio [8086:a2a3] 00:1f.4 SMBus: Intel Corporation 200 Series/Z370 Chipset Family SMBus Controller It's 00.1f.3 I try to add. I get this error: Syslog: Feb 14 02:51:16 Tserve kernel: vfio-pci 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none Feb 14 02:51:16 Tserve kernel: br0: port 2(vnet0) entered blocking state Feb 14 02:51:16 Tserve kernel: br0: port 2(vnet0) entered disabled state Feb 14 02:51:16 Tserve kernel: device vnet0 entered promiscuous mode Feb 14 02:51:16 Tserve kernel: br0: port 2(vnet0) entered blocking state Feb 14 02:51:16 Tserve kernel: br0: port 2(vnet0) entered forwarding state Feb 14 02:51:17 Tserve kernel: vfio_ecap_init: 0000:01:00.0 hiding ecap 0x19@0x900 Feb 14 02:51:17 Tserve kernel: br0: port 2(vnet0) entered disabled state Feb 14 02:51:17 Tserve kernel: device vnet0 left promiscuous mode Feb 14 02:51:17 Tserve kernel: br0: port 2(vnet0) entered disabled state Feb 14 02:51:18 Tserve kernel: vfio-pci 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none PCIe ACS override is set to Both and split the GPU to different IOMMU groups. But not for the Intel group. I've tried "VFIO allow unsafe interrupts" both No and Yes, but I get the same error. I can't find any way to add the other things in that IOMMU group to the VM. Is there a way to get the Intel/motherboard soundcard to work, or is it impossible due to the structure of the motherboard?

Tried another bulk now, and changing the "minimum free space" did the trick. Thank you!

Thank you for the answer. It's set to the default 0KB. Changing it to 10GB now, will see if that helps for the next bulk.

Transferring quite a bit of data to my unraid server at the moment, from a Windows machine using SMB. The share is set to "Yes" for use cache. The cache is at 250gb, a lot less than the amount of data I'm transferring. The problem: When the cache-disk is full, the transfer stops, claiming the disk is full. All I have to do is to press resume and the transfer continues, filling up the array disks instead. But I feel it shouldn't stop at all, that unraid should move over to the array automatically. The description for the cache settings when creating a share indicates that: Was thinking that maybe Prefer is the correct alternative to fix this, but it has the exact same function as yes in the description: Do you know how I fix this? If there isn't a fix, just a thing with unraid, is it only with SMB, or is it the same with all protocols, dockers, and apps? When cache-disk is full, they all stop and have to be manually resumed?

Perfecto! Thanks!

Okey, thanks. Guess that's the best option. It still works in windows though. Just in a weird way. Do you know if there's any full sector check for unraid? Running one windows now with HDDScan, but it looks to take days, so I'd rather do it on unraid when it's up as it will be running 24/7. Would be interesting to see if there are bad sectors, or if my adata disk us just weird - and perhaps reuse it for some other more unimportant stuff.