hawihoney

I'm not sure if my failing drive was a parity drive before. But as it's SMART values are ok I bet it was. I don't swap data drives if SMART values are ok. But I swap parity drives with perfect SMART values if I need a bigger one.

Yes, wipefs did the trick. Double check that this drive is not part of the array. Double check the correct device name. Issue "wipefs /dev/xxx" (replace xxx by correct device) without any additional parameters to check the current FS. This step is not necessary. It's just to check if it gives a non-unRAID-FS output. Finally "wipefs --all /dev/xxx" to wipe the disk. Lasts a second or so. After that I could attach the disk immediately to UD.

I tried this twice now. I doesn't work for me. After a reboot, and when starting the first rsync between both machines, I'm always welcomed by this message: root@Tower2:~# rsync -avPX --delete-during --protect-args -e ssh "[email protected]:/mnt/user/xyz/" /mnt/user/xyz/ The authenticity of host '192.168.178.35 (192.168.178.35)' can't be established. ECDSA key fingerprint is SHA256:SSFXwWXedKMmxBao0vvheifFEfIoiiQl5rtfPuZ8x3w. Are you sure you want to continue connecting (yes/no)? Here are my steps I did twice: On Tower: ---------- # ssh-keygen -t rsa -b 2048 -f /boot/config/ssh/tower_key Generating public/private rsa key pair. Enter passphrase (empty for no passphrase): [press enter here] Enter same passphrase again: [press enter here] On Tower2: ----------- # ssh-keygen -t rsa -b 2048 -f /boot/config/ssh/tower2_key Generating public/private rsa key pair. Enter passphrase (empty for no passphrase): [press enter here] Enter same passphrase again: [press enter here] On Tower: ---------- # scp /boot/config/ssh/tower_key.pub root@tower2:/boot/config/ssh On Tower2: ----------- # scp /boot/config/ssh/tower2_key.pub root@tower:/boot/config/ssh On Tower: ---------- mkdir -p /root/.ssh cp /boot/config/ssh/tower_key /root/.ssh/id_rsa cat /boot/config/ssh/tower2_key.pub > /root/.ssh/authorized_keys chmod g-rwx,o-rwx -R /root/.ssh On Tower2: ----------- mkdir -p /root/.ssh cp /boot/config/ssh/tower2_key /root/.ssh/id_rsa cat /boot/config/ssh/tower_key.pub > /root/.ssh/authorized_keys chmod g-rwx,o-rwx -R /root/.ssh Add to /boot/config/go on Tower: ----------------------------------- mkdir -p /root/.ssh cp /boot/config/ssh/tower_key /root/.ssh/id_rsa cat /boot/config/ssh/tower2_key.pub > /root/.ssh/authorized_keys chmod g-rwx,o-rwx -R /root/.ssh Add to /boot/config/go on Tower2: ------------------------------------ mkdir -p /root/.ssh cp /boot/config/ssh/tower2_key /root/.ssh/id_rsa cat /boot/config/ssh/tower_key.pub > /root/.ssh/authorized_keys chmod g-rwx,o-rwx -R /root/.ssh I really would like to get this running. Many thanks in advance.

I don't know if this is a Docker or Unassigned Devices thing. I do receive permission denied for all sorts of files and directories. If I look at directories and files from within docker User "abc" and Group "abc" have permission. Outside the docker container it's "nobody:users". The docker container is running as "99:100" and all directories but "/config" are set to "Slave/RW". What else do I need to do? Here's docker view vs. unRAID view: root@Tower:/mnt/disks/UA01/nzbget# docker exec -it nzbget ls -lisa /nzbget/intermediate total 12 4297828992 0 drwxrwxrwx 3 abc abc 73 Mar 26 10:38 . 99 0 drwxrwxrwx 7 abc abc 124 Mar 27 07:59 .. 2147483745 12 drwxrwxrwx 2 abc abc 8192 Mar 26 11:03 'Test.#17489' root@Tower:/mnt/disks/UA01/nzbget# ls -lisa /mnt/disks/UA01/nzbget/intermediate/ total 12 4297828992 0 drwxrwxrwx 3 nobody users 73 Mar 26 10:38 ./ 99 0 drwxrwxrwx 7 nobody users 124 Mar 27 07:59 ../ 2147483745 12 drwxrwxrwx 2 nobody users 8192 Mar 26 11:03 Test.#17489/ Attached you'll find the important settings. If you need diagnostics I can attach it too. Any help is highly appreciated.

Thanks for posting. I can't see custom br0 on my docker details - it's custom eth0 only. What do I need to do?

This is the output of wipefs. DOS? On a 3TB drive? Last time I used DOS was when using 1,44MB diskettes. Really weird. Linux 4.14.26-unRAID. root@Tower:~# wipefs /dev/sdg DEVICE OFFSET TYPE UUID LABEL sdg 0x1fe dos root@Tower:~#

This drive was part of an unRAID machine. I did replace it with a 6TB. Don't know what filesyste it had (ReiserFS, XFS, BTRFS or parity). Do I issue wipefs for the partition sdg1 or the drive sdg? Thanks for helping.

Ah, sorry. Diagnostics attached. tower-diagnostics-20180325-1649.zip

Today is the first time to use Unassigned Devices. I took an old drive and did insert it into my unRAID box. I do see the drive in UA and UA reports about an existing partition. Whenever I try to remove that partition, it reports success but the partition is still there. I would like to format that partition in XFS and share it's content. What should I do? Thanks in advance. ***Edit*** Destructive mode was and is enabled. Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] 5860533168 512-byte logical blocks: (3.00 TB/2.73 TiB) Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] 4096-byte physical blocks Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] Write Protect is off Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] Mode Sense: 9b 00 10 08 Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] Write cache: enabled, read cache: enabled, supports DPO and FUA Mar 25 15:07:59 Tower kernel: sdg: sdg1 Mar 25 15:07:59 Tower kernel: sd 11:0:5:0: [sdg] Attached SCSI disk Mar 25 15:08:29 Tower emhttpd: WDC_WD30EZRX-00MMMB0_WD-WCA********* (sdg) 512 5860533168 Mar 25 15:09:36 Tower unassigned.devices: Removing partition '1' from disk '/dev/sdg'. Mar 25 15:09:48 Tower unassigned.devices: Removing partition '1' from disk '/dev/sdg'. Mar 25 15:10:44 Tower unassigned.devices: Removing partition '1' from disk '/dev/sdg'.

Thanks for the answer. Mail is on it's way.

I do have five old (~10 years) License keys and original LimeTech USB devices from the days I bought two machines from LimeTechs website. Some of the USB devices are under 512 MB in size. So today I tried to replace the first of my Pro licenses to a new USB device. I did as requested on that page: https://lime-technology.com/replace-key/ The original key file (here: Pro5.key) from the old device was the only keyfile on the new device. The machine came up and complained about to many devices. So I clicked on "Replace key" on Tools-> Registration and did enter my E-Mail address. The following screen appeared (see attachment). No mail arrived, No new keyfile arrived. What can I do? When I bought the two machines I ordered a second USB device and license always. Perhaps I did replace a device in the past, but I don't know which. If I did replace, it was many years ago.

Thanks. The announcement of USB creator tool did include the 'make bootable' feature. Has this been removed?

Whenever I try to work with the USB creator tool, it does not do what I expect. What's wrong with my approach? ~10 years ago I bought two LimeTech server from the website. From these days I own five Pro license and USB devices. Today I tried to bring them all to the latest content. Not one single device worked after preparation with the USB creator tool. What I did: - Stop both unRAID server - On each server click on the Flash device on the Main page of the unRAID GUI - On each server click backup and wait for completion - Copy the backup files of both server to a safe place - Shutdown both server - Insert the USB devices in a Windows PC - In Windows Explorer format USB devices to FAT32 (label UNRAID) - Start USB creator tool (yes to Windows Admin request) for all USB devices - In USB creator tool click on local ZIP file - Start creation and wait for successful completion - Double check that the five license files do exist in the config directory (Pro1.key, Pro2.key, ...) - Unmount USB devices from Windows PC - Remove USB devices from Windows PC and insert into matching unRAID server (two server, two different flash contents) - Start unRAID server SYSLINUX 6.03 EDD 2014-10-06 Copyright (C) 1994-2014 H. Peter Anvin et al WARNING: No configuration file found boot: This happens for all USB devices I did create with USB creator tool today. Comparing both devices (the original one and the one created today) with Total Commander look identical. Shutdown server, remove newly created USB devices and put old USB device back into server works as expected. What is this? Why doesn't it work for me?

Combining different folders of two different ZIP files is not what I call a user friendly restore process. There's a backup button for unRAID OS, that will include all user settings, on the details page of the flash device. Would be better on the Tools page but this is a different story. So far so good. There's a USB creator tool to restore a backup of the unRAID OS including all user settings on a flash device. So far so good. Now this USB creator tool needs 1.) an internet connection and 2.) access to the latest unRAID files (on AWS). There's no chance to restore a locally existing backup archive without 1.) an internet connection and 2.) without access to that AWS cloud service. I do have the backup archive but I can't install it because the files on AWS are not there. So, while USB creator tool is broken and refuses to restore my existing backup files, I need to prepare the flash manually. I do have the locally existing backup but I can't create the flash without 1.) an internet connection or 2.) without access to that cloud service and the unRAID OS base files. And on top of that I need to decide what unRAID OS version is running, download the corresponding unRAID base files and combine parts of two ZIP files to create my flash. Here in Germany we call this "frickeln" - in english "fiddle/tinker" (c) Google. All this would not be neccessary if: 1.) Flash backup would include the syslinux folder in the backup archive (for manual restore) 2.) USB creator tool would allow restore of an existing backup without an internet connection and access to files that I don't need for a restore.

I come to the conclusion that the whole process is not bulletproof. I took the latest flash backup, did format the flash and copied the whole contents of the flash backup ZIP onto the flash. Right click on make_bootable.bat with Administrator Rights --> H:\syslinux\syslinux.exe missing. The ZIP of the flash does not contain the syslinux folder. A manual restore (without the non-functional USB creator tool) is not possible. So I need the flash backup ZIP AND the corresponding ZIP of the matching unRAID release to combine a USB manually. My feature request: 1.) Allow the restore with USB creator tool without the outbound connection. A user simply wants to put a flash backup ZIP on the flash. He does not want to fetch the latest release from the website. 2.) Put all required files for a restore in the flash backup ZIP. Just my 0.02 cents.

Seems that I do have bad luck today with unRAID. My flash was corrupt after the update to 6.5. After that I tried to put a backup on that flash but the USB Creator tool gives an error (please see screenshot attached). What can I do?

Updating the package from 5.24 to the most current one sounds like a good idea. With the package included in the Nerd-Pack even CPAN itself needs an update. Is looking for newer packages a manual process? Thanks a lot.

I'm looking for some guidance regarding the Perl package. It's version perl-5.24.0-x86_64-1.txz is included in this Nerd plugin. Whenever I restart my servers hundreth of CPAN modules need an upgrade. So I need to issue "cpan -upgrade". This call does a ton of updates. Most of the things end successfully, some fail. In addition I need to install two additional modules (e.g. File::Listing). I would like to avoid that upgrade step after a reboot. It's lengthy (something around 15 minutes), it's faulty and it needs some keystrokes in the middle of the upgrade. Is there a better way to get Perl up and running after a reboot? With up-to-date modules? With additional modules? How do you work with Perl after a server reboot? Many thanks in advance.

Thanks again for your outstanding support.

In the MakeMKV support thread you told me how to call MakeMKV from within a Perl script: docker exec MakeMKV env HOME=/config /opt/makemkv/bin/makemkvcon [args...] What's the correspronding call for the tool mkvpropedit in the MKVToolNix docker? The following doesn't work: docker exec MKVToolNix env HOME=/config /opt/mkvtoolnix/bin/mkvpropedit [args...] Many thanks in advance.

You are my hero. I usually would try to add a r/w path "Mnt" with "/mnt=/mnt" but I bet this is not possible?

Thanks a lot, again. Copy/Paste worked. Perfect. Calling makemkvcon within docker from a perl 1-liner works. Perfect. My mapping is docker=/storage equals perl=/mnt. The perl script populates e.g. "/mnt/disk1/folder". So I do have to replace /mnt by /storage within the makemkvcon command giving e.g. "/storage/disk1/folder". Right? I can see a light at the end. Thanks.

One week later. The MakeMKV docker is working now. Two additional questions if it's allowed: 1. I need to set the options on the expert tab of MakeMKVs GUI. The entryfield on that tab does not accept pasted values. Is there any way to paste data into that entryfield or do I need to edit that manually (the sel values and codecs)? BTW, are changes here persistant? 2.) I did install the MakeMKV docker because I did write a cool perl script to walk thru my disks and folders to convert all my ISOs to MKVs automatically. On Windows, over LAN, it's working perfect. From within the perl script I call makemkvcon. This executable is part of the MakeMKV docker. I'm pulling my hairs out because I don't know how to call makemkvcon, that's part of the docker, from within my perl script. Perl is installed with the Nerd package. So I want to run a perl script, call system(makemkvcon..) from within the script and makemkvcon is part of the docker. And makemkvcon should respect the license key and my settings on the MakeMKV expert options. Is this possible? As an alternative I would compile MakeMKV but it references so many libs that I don't find on slack. Any help is highly appreciated.

Ah, thanks. It's starting now. I still need to know why I do need an optical drive to work with ISO or MKV files. When starting without the extra parameter "--device /dev/sr0" the tools (e.g. makemkvcon) complain about a missing device. When starting with the extra parameter "--device /dev/sr0" the docker fails to start because of a missing device. Faking with "mkdir /dev/sr0" did not help.

What is CA? I simply gave up. This docker stuff is way beyond my knowledge. I was looking for an easy way to install makemkv on my systems. Docker is not the way I can go. What I did (took me some hours): docker run --name=makemkv -e MAKEMKV_KEY=thekey -v /docker/appdata/makemkv:/config:rw -v /mnt:/storage:rw -v /mnt:/output:rw jlesage/makemkv Then after calling: docker exec makemkv /opt/makemkv/bin/makemkvcon info /mnt/disk3/Series/ABC/S01E01.iso I was told that an optical drive is missing. I don't have optical drives in my machines. So I gave up.