maciekish

Hi, I disabled all VMs and Dockers and the system has been up for 5 days. I have run memtest and prime95 for 29 hours without issues. I am currently reenabling a docker or vm every 48 hours to until it crashes. If it doesnt help i will connect a monitor and keyboard but it is very difficult so i am leaving it for later.

The server hung again today. Can someone please advise? I am completely out of ideas. HANG tracer-diagnostics-20180724-0733.zip

The box keeps hanging at random every 24-48 hours. Yesterday it hung and i walked over with a keyboard and monitor. Monitor didnt come on, but i could switch Num Lock on and off, and if i switched tty with ctrl+alt+f1 and f2 the Num Lock switched on and off (one tty had on and the other off?). Logging in and rebooting blindly didnt work. Had to hard-reset. Today it hung again as i started a Plex sync. I have removed all USB devices (UPS and FR24 receiver) and disabled the two Ubuntu VMs now. Logs attached. I ran memtest86 for an hour a couple days ago, no errors reported. This is driving me nuts! FCPsyslog_tail.txt tracer-diagnostics-20180721-1724.zip

Hi how do i disable the continuous flash logging of syslog and diagnostics? I enabled it a couple days ago and cant find how to disable it any more.

It is controlled by swappiness. I guess you could create a userscript on startup that sets the swappiness. https://askubuntu.com/questions/103915/how-do-i-configure-swappiness I suggest the plugin author adds an option for this in the plugin. sysctl vm.swappiness=[value]

Hi, For the second day in a row my unRAID server has hung. Yesterday i could SSH in but i could barely run anything. When i tried top or htop it would just chug until i pressed ctrl+c to cancel. Today ssh didnt work either, it just timed out. I have a couple dockers and two Ubuntu virtual machines. All the hardware except a few disks are brand new and the system worked fine for 1 or 2 months until yesterday. After reboot everything is fine. I havn't made any changes i can remember and the system isnt under any particularly high load that i am aware of. Any ideas what could be wrong? I was only able to capture diagnostics after restart unfortunately. tracer-diagnostics-20180705-1653.zip

Obviously i would check with the device identifier but what am i looking for and where please?

Where in the diagnostics can i check it?

tracer-diagnostics-20180619-2017.zip

Hi, I just added a second SSD to my system and selected it as Cache 2. unRAID says the "device is part of cache pool" but there is no activity on it and im not sure this is correct. I want the new SSD to mirror the other. root@tracer:~# btrfs filesystem show Label: none uuid: f8232945-44f6-415a-a32b-72f79cf16d6a Total devices 1 FS bytes used 241.77GiB devid 1 size 931.51GiB used 330.02GiB path /dev/sdc1 Label: none uuid: 08254a30-721e-4608-a81c-cb94a1632247 Total devices 1 FS bytes used 14.22GiB devid 1 size 30.00GiB used 18.52GiB path /dev/loop2 Label: none uuid: 5ea50f9e-b636-4219-8256-48303c232da9 Total devices 1 FS bytes used 236.00KiB devid 1 size 1.00GiB used 126.38MiB path /dev/loop3 root@tracer:~# btrfs fi df /mnt/cache Data, single: total=328.01GiB, used=241.25GiB System, single: total=4.00MiB, used=64.00KiB Metadata, single: total=2.01GiB, used=541.05MiB GlobalReserve, single: total=259.77MiB, used=0.00B

Hi, I have 3x Seagate IronWolf 10Tb drives with one as parity and a 1Tb SSD as cache. Once the cache is full and the IronWolf internal 256Mb cache is full write speeds drop to 40Mb/s from 115Mb/s. I realize unRAID has to read for each write but shoud it really be this slow or am i doing it wrong? LSI 9201-16i controller, same speeds on mainboard controller as well.

Hi, I have one network adapter in my unRAID box, eth0. I have set up Caddy with its own ip on eth0 in Docker as a reverse proxy for multiple services and subdomains. I want a reverse proxy for the unRAID interface as well, but i get 502 Bad Gateway. If i do docker exec -it Caddy bash and then try to ping the host IP it times out. Seems like the traffic is blocked or not properly routed. How do i fix this? Network is 10.0.0.0/22 Host is 10.0.1.0 on eth0 Caddy is 10.0.1.12 on eth0 in Docker I'm able to ping other devies in the network for example the router at 10.0.0.1.

I’m planning to flash it to IT mode and not use onboard RAID so i guess i’ll be fine then. Thank you for your time!

Hi, is it possible to move disks between controllers without losing data? I want to put my server together now but my LSI controller won’t arrive for a while longer. Can i put a couple disks on the mainboard now and just move them to the co troller when i get it? Its an ASRock manboard with a standard Intel controller and im getting an LSI 9201-16i.

Thanks for CA Backup, might be all i need. Is there anything similar for VMs? And yes, i know its not a backup substitute, everything important goes to Glacier on my current setup and will on unRAID as well. Would it make sense to set the appdata share to "only" cache in this case then?

Well thats the „problem”. I only have one cache SSD for now so i want cache „yes” not „prefer”. The question is whether the mover will mess up the permissions if i do this. Reason being all appdata would be unprotected with cache „prefer” otherwise.

Hi, i want to run Dockers and VMs on a cache drive. The drive would be used as regular cache drive as well with the appdata folder set to "Yes" for cache which means apps write to SSD first and it is moved to spinning disks at night for redundancy. Now i read somewhere but dont remember where that this "mover" messes up permissions for the appdata folder and i will have problems with my dockers after this runs. Is this still true or was this old info? Don’t understand how this would be a problem since appdata woult be accessed via /mnt/user/appdata which is a FUSE? Is it true that this would be a problem and i need another SSD for a cache pool if i want redundancy?

Hi, I currently have a Qnap with Container Station which does not allow easy upgrading of images. Basically i have to delete the container (with settings), delete the image, redownload and reconfigure the docker. Stupid. How is this handled on unRAID? Can i upgrade an image while keeping my settings? Qnap hides everything away. User friendly. And unusable.

I have a discrete GTX 980M not integrated graphics but it sounds like one GPU doesnt cut it. Thanks for the explanation.

Hi, I just discovered unRAID and I'd like to know if its possible to use it like Qubes to run multiple systems at a time from the "console". Ie does it have a window manager like Qubes or will i have to reassign my GPU every time i want to switch to a different machine? I have a very powerful laptop (desktop CPU with VT-d and VT-x, 32GB of ram etc) so performance is not an issue. I want to be able to install Windows 10 and basically "alt-tab" to Linux or perhaps Mac OS + use all the other unRAID features I've read about. Bonus question: Does each VM running require its own GPU or can i assign my GPU to Windows but still have another OS running in software mode and freely switch the GPU between the machines?