Struggling with Raid-5 on Ubuntu Server 18.04? This setup can be tricky.
Struggling with Raid-5 on Ubuntu Server 18.04? This setup can be tricky.
Most of my mistakes come from not understanding what to search for online, relying on guides, and then applying them without fully grasping the details.
The pool is functioning properly as expected. The operation executed was a GET request, retrieving details based on the variables surrounding it. ZFS was implemented before and will continue after your request, displaying all relevant zfs information. Seeing 'new-pool' confirms the mountpoint exists, indicating the pool was mounted at creation—this is convenient. It simplifies future tasks. You can access it using: cd /new-pool. As a logged-in user (jamess7995), verify your rights to create, modify, and remove files in /new-pool. Use ls -l /new-pool to inspect ownership and permissions. For testing access, run commands like: cd /new-pool touch file1 echo 'echo execution works!' >> file1 chmod 744 file1 ./file1 rm file1 mkdir dir1 rmdir dir1 These steps help confirm functionality across file creation, editing, permission setting, execution, and directory management.
I executed those commands, but the only one that failed was the echo command. For some, I had to use the sudo prefix. Commands weren’t working due to caps lock, and it seems the keyboard’s LED wasn’t functioning properly. That’s why there’s a white block in the middle.
Linux treats case differently. R and R are distinct, as are K and K. Given the permission problems I’ve faced, it seems the owner is likely root. You shouldn’t rely on sudo for everything—once you enable the SMB share for Windows, you’ll lose access in the GUI, making it impossible to use the network share at all.
Although you can change the owner of the directory I've had people yell at me saying it isn't the proper way to do it so here's a quote from what is supposedly the proper way to setup a SMB share: If not all of this makes sense it's saying to create datasets inside /new-pool. Datasets behave similar to directories but give you a lot more control as far as what you can do with them. These datasets can then be assigned specific users (such as jamess7995) and hold specific datatypes for organizational purposes. They can each be their own SMB share if you really wanted as well. #3 is talking about how permissions will be handled on the dataset(s) itself. There are options including UNIX | Mac | Windows. I don't fully understand it myself but I believe setting it to Windows (restricted) has something to do with giving the Windows clients full control over setting permissions which would be preferred for your application. Now the specific commands to accomplish this I'm not sure of since I used the FreeNAS WebUI. Gonna have to do some more research but I was able to give you a couple arrows at least .
Noted, the idea involves organizing files into folders like a directory, keeping ownership in mind. My aim is to set up a parent folder, a media folder, and upload from my Windows PC to the media folder via SMB. I want Plex on the server to let PS4 or other devices access the data. The printer behaves oddly with computers, but I’m open to scanning to a server if it works. I prefer using a custom username and password for SMB, though that’s just my preference.
I think there might be a more straightforward solution. My goal is a samba share that works across multiple usernames and passwords, using the three hard drives and the flash drive as the operating system when needed—just like in the example above. I’d also like to run a DLNA media server so Plex can be used easily. If this can be simplified with another OS, I’m open to hearing your thoughts.
GUI and web interfaces simplify tasks for newcomers, but diving straight into the command line offers greater productivity and precision. For optimal control, using CLI is recommended. Currently, you're about a quarter of the way through. You already have the pool established. Next steps include building your datasets in subfolders, configuring SMB shares, adding users and assigning them to datasets. Once printers support SMB output, that should also be feasible. Research suggests creating datasets is straightforward: use zfs create with appropriate mountpoints and dataset names. For example, zfs create -o mountpoint=/mnt/binaries datastore/binaries or zfs create -o mountpoint=/mnt/homes datastore/homes. Your commands would resemble: zfs create -o mountpoint=/mnt/jamess7995 new-pool/jamess7995 zfs create -o mountpoint=/mnt/parents new-pool/parents zfs create -o mountpoint=/mnt/home new-pool/home zfs create -o mountpoint=/mnt/backups datastore/backups. After setting up the mount points, accessing datasets via /mnt is recommended instead of the temporary directories. All data written to these datasets will be stored physically on your pool. Then you can configure SMB/CIFS shares and adjust permissions according to user needs, using group settings for access control.