I know... It's been a long time. Work and life has kept me very busy. I need to get back here more frequently. I really enjoy trying to help others out.
Thanks for the reply :) I did get this...
Type: Posts; User: rubylaser; Keyword(s):
I know... It's been a long time. Work and life has kept me very busy. I need to get back here more frequently. I really enjoy trying to help others out.
Thanks for the reply :) I did get this...
It's been years since I've managed to screw up a server like this, but today was my lucky day. I picked up a cheap 1050GTX to try to use GPU decoding/encoding for my Plex Docker container. I...
I just stumbled across this the other day. This looks like a nice, free, open source way to automate the transcoding of disks. You could even extend the scripts to automatically scan for the new...
With a spanned LVM volume on top of a bunch of disks without parity (like mdadm), you will lose all of your data in the in the event of a failure. I find MergerFS to be perfect for what you are...
The main advice is to make sure you are first working on the correct disks, so you don't overwrite the wrong one. Here are some good directions (you can skip the Knoppix part and really the smart...
This is tough. Right now, you have one disk that is so out of date because it's been removed in the past that you can't really use it for reassembly and one that is not working well (read errors). ...
Yes, you should add /dev/md0, sorry, I typed that on my phone yesterday :) Also, /dev/sdc has errors, but they appear to be the result of UDMA errors which is normally the result of a bad SATA...
This sounds like it's probably a permissions issue (readable only by the root user and not the plex user). Try chowning the files or chmoding them to more lenient permissions.
Did you already run the --create --assume-clean command (--create should be the absolute last thing you try). If, not, please don't do that yet. Your previous attempt at --assemble did not use the...
resolv.conf is generated automatically on reboot. If that's all you did, just re-create it, and restart the interface. Do you have a dns-nameserver line in your /etc/network/interfaces?
I'd start...
I did answer your question on linuxserver.io, but I'll paste my answer here for you too.
This is old, but it does seem related to apcupsd. FYI, I use NUT to manage my UPS. This suggests disabling...
For anyone struggling with disconnects with mhddfs, I'd suggest to tryout mergerfs. I have switched away from both AuFS and mhddfs to mergerfs and it works great. It is actively developed, does not...
Another option is to use gddrescue on the failing disk to recover to a new disk. This will try harder to recover the data on the failing disk. If the recovery works, you can try to add the new disk...
These are all fine suggestions. But I would keep your OS disks separate from your bulk data disks (I'd run the OS on a couple inexpensive 120GB SSDs in RAID1). For bulk home media, I switched over...
Darko's directions are correct. After you update your mdadm.conf file, you should also update your initramfs like this.
sudo update-initramfs -u
Assuming corruption to be in a certain order doesn't make sense unless you have verified that this is the case. With RAID5, because of striping, if you have corruption, the data is corrupt period,...
I would assume that rsync re-scanning and syncing is updating a timestamp or metadata on previously synced file and that is what is causing the restored message.
mlimit should be honored if a non-root user is performing the write. Unless you get to the point where all disks in the pool have less space than the threshold, then the disk with the most free space...
You will have to manually move files off the data disks that are too full. Move about 20GB off /mnt/6XW0QL54d01, /mnt/6YD05MN6d02, /mnt/9XW06E43d04, /mnt/6XW0V05Fd05 onto one of your disks with lots...
How full are your 1.5TB data disks? I would assume that one or many of them are completely full, and there isn't enough room on the 1.5TB parity disk to hold all the parity (the 2TB disk should be...
You need to zero the superblock on the disks that were in the array, or as long as mdadm is installed, it will try to re-assemble your array.
mdadm --stop /dev/md0
mdadm --zero-superblock...
If you wanted to give it a try. I'm currently back to using AUFS. I've compiled it myself to support hnotify (I have a reference to how to do it in my SnapRAID tutorial in my signature). So far,...
Yes, the format for /etc/fstab needs to look like this.
mhddfs#/media/disk1,/media/disk2,/media/disk3,/media/disk4 /storage fuse...
SnapRAID handles this situation very well (I've upsized my parity drives 2 times now). All you need to do is rsync your current parity file to your formatted, new larger disk. Once done, you replace...
Yes it does honor the mlimit even for root. You probably want to set a larger mlimit than 5GB with many small files.