I'd best carefully reread the whole guide to see what else I've missed
If you plan to stick with OMV, it's worth a full read through. The idea is to get started. Later on, this -> web site has a few additional doc's that might be of interest to you.
OMV really impresses.
I agree. OMV is very good and it appeals to a wide variety of skill levels, from PC hobbyists up to and including business admin's. I was using Windows Home Server which MS abandoned. Looking for something with the same functionality brought me here where I found a world of server add-ons and actual (real) data protection. The bonus is, OMV is so light on resources it will run from a Raspberry Pi 4 nicely (a low cost single board computer). That makes for cheap backup.
Frankly, I wouldn't have paid the price for WHS upgrades even if they'd have continued to support it. It's far easier to throw $10 USD at OMV, to help keep it going, where I can have my cake and eat it too. With Dockers and other server add-on's, advanced tech is in the hands of PC enthusiasts at no cost (except for a modest donation here and there).
Coming from a mainframe world of 20 years ago and with a reasonable level of experience supporting PC hardware and using Microsoft software products back to MS-Dos would either of you care to recommend a good Linux primer and a distro to experiment with?
I'm in roughly the same age group you are. Realize the following are my personal preferences but, when I started to look, I was looking from something that was roughly similar to a Windows Desktop.
In general, I'd go with a desktop distro based on Debian or Ubuntu (Ubuntu is, ultimately, based on Debian). My preference for a Desktop was KDE because I liked the native applications. KDE app's are nicely polished. However, there's nothing wrong with XFCE for hardware that's older and a bit under powered. XFCE is similar to Windows 7.
You can get a sense of distro popularity here -> Distrowatch.com (Look at the lower right, "Page Hits", column.)
For distro's that approximate Windows, this is worth a look. -> Linux for Windows Users
I think the best Linux primer is just "do it". Spin up a Linux desktop and, when you run into issues, search the net for answers. There's a lot of free references and tutorials on-line.
The next bit explains how to set up a scheduled job – but doesn’t advise how to get the uuid into the command – and you can’t copy the content of the mount point field as far as I can see. You say “it is easier if you use auto completion, that means, type the first letters or numbers and the press the TAB key”.
If you navigate to the drive, using -> WinSCP (also in the guide), you can copy and paste the drive location into your command line (complete with UUID).
(The following is not a UUID example, but the same concept applies.) Highlight the Address line, Ctrl+c (copy) and Ctrl+v (paste) and paste it into the command line. Add an additional / at the end of the drive location and you're good to go for the drive entry.
The "Drive by UUID" was a midstream change in OMV5 and there didn't seem to be an "elegant" way to patch the existing guide.
Sorry the guide doesn't quite go into enough detail for you but some conventions (like Enter / Return being the same thing) have to be assumed. The guide is close to 90 standard pages already.
BTW: The guide is intended to be an "OMV New User Guide". Guides and books for new Linux users, general computing conventions and accuracy in typing are another type of reference altogether.
Letsencrypt won't do anything for you. The only benefit in using Letsencrypt is if your hard drives are physically stolen. (Highly unlikely.) The use case for Letsencrypt is for Laptops or business servers where the risk of theft is a possibility.
If you don't open server ports at your router, which exposes your server to the Internet, you'll have very good server security.
I hope, that as some point you remove the additional code, that blocks using USB(3) disks as RAID.
He didn't write the code that blocks RAID. That limitation is part of the main project.
There's a reason why this restriction is in place. It's a bad idea, it fails a lot, and most users run without data backup. That's a sure formula for users to return to the forum, asking for support for fixing a broken array that, in many cases, can't be fixed.
If you really want to run RAID, it can be set up on the command line. Thereafter, the array will show up and can be managed in the GUI. But, if you do this, you've been warned.
While it's possible to upgrade, I'm not a fan of it. If an existing install has an issue, that problem is carried over into the next version. For this reason alone, a fresh build with installing a new version, every 4 or 5 years, is a good idea.
Setting the above aside; the hardware OMV is installed on varies widely, along with numerous user customizations and software add-on's. Accordingly, the script and upgrade process is not "guaranteed". Further, trying to reverse engineer what might have went wrong in an upgrade can take a lot of time and the exact reason for a problem might never be discovered.
For these reasons and more, I'd recommend building from scratch. This assures a clean install. Your call.
Perhaps others may have useful input regarding your specific error(s).
You can check the flashmemory plugin with:
folder2ram -status (and)
Look for "successful" in the various /var folders.
If it's not running (in OMV5) it can be installed with the following one line script command:
wget -O - <a href="https://github.com/OpenMediaVault-Plugin-Developers/installScript/raw/master/install">https://github.com/OpenMediaVault-Plugin-Developers/installScript/raw/master/install</a> | sudo bash
Because the drive needs to be mounted manually first to rebuild the RAID, I need access to a CLI.
This left me on the WebGui not able to get on the CLI in the first place.
Is there a Plugin or so that allows CLI access from the webpage?
Or do I need to SSH remotely?
SSH in from a LAN client. This will get you started. -> PuTTY .
In general it would be a nice feature in OMV to handle a broken raid completely remotly via WebGui (being a Linux Rookie this might be just a nice User-Wish). This makes the Likely case more easy for unexperienced users to rebuild their NAS RAID...
OMV can handle RAID recovery from the GUI but handling the hardware end of it, in a VM, is not the same.
On physical hardware:
- Remove the failed disk (be sure you remove the correct disk) and install a new one.
- Boot up.
- Wipe the new disk in Storage, Disks
- In Storage, RAID Management, click the Recover Button.
- Add the new disk to the array.
**EDIT: You can even get past the physical drive swap "if" you have an extra drive in the server. In that case, it can all be done by remote.**
I'd cut the heatsink (just metal and low cost) before I'd cut the motherboard connector. It might be hard to get the NVME's thermal pad to release. In any case, bad news.
There are NMVE to USB adapters but you wouldn't get the phenomenal speed of the NMVE when it's plugged into USB. Even if the NVME is fine, there's the socket on the motherboard to consider. (And there's the adapter order delay time to consider as well. An RMA is usually limited to 30 days.)
On the other side of it, I'm assuming that the thumbdrive is working fine. You really don't need fast boot media unless you're doing virtualization or, maybe, storing more than one or two dockers on the boot drive.
I've never had any trouble with using Thumbdrives to boot in X86, as long as the Flashmemory Plugin is used and it's quality drive like SanDisk. The story is "roughly" the same with SD-cards in SBC's when buying a quality item (SanDisk or Samsung). It does "seem" as if SD-cards are more susceptible to dying young (junk cards don't last), but there are no worries if backup is maintained.
On the other hand, there's nothing wrong with an SSD.
Don't forget to install the Flashmemory plugin. (Required for Thumbdrives.) Get that done as soon as your build is complete.
There are reference links above.
Return the NVME. Try building on a thumbdrive as test and, maybe, stay with it if it works for you.
In my case, they've been working fine.
I wouldn't be overly confident yet. I'm not very familiar with SSD's in general (I do have one). However, the inconsistency of SMART data returns, when queried, would still be a concern to me.
One could make an image of the install when everything is up and running, save that on some safe location. That way you just need to burn that image and it will be restored.
I do the above with dated image files AND have I a cloned thumbdrive ready for my main server, sitting on top of the case. That allows some rollback depth in backup and, with a backup boot device built and in hand, recovery can be done is as little as 3 minutes.
I'm going to give the timeshift package a closer look, this fall. For layered OS backup, that seems like the optimum approach.
There's a script that will upgrade from 4 to 5 but there may an issue (or two) when running it. It's better, in my opinion, to do a clean install versus rolling an existing but unknown problem, into a new version.