It's now resolved.
Came here this morning to see why it was broken, and found the teamviewer repo error....
Wondered why, as I don't have it enabled, and then found it was claiming a repo signing error, but it's actually 403 rejecting everyone who tries to connect.
I presume they've broken it nicely, so I've temp commented it out, and (amusingly) have nothing to update.
Hopefully it's back in 24 hours or so...
I've got the firmware image also (just omitted it as it wasn't headers/kernel).
They all show up on the grub menu - but that's where I noticed - after the crash I had, the boot menu was "large"
Well, it's certainly adding them - I didn't know if it had a limit (I'll soon find out if it hits 11). I know votdev has commented before that it's an owners responsibility normally to remove them in case stuff breaks
apt-get --simulation autoremove shows nothing to remove.
dpkg -l | grep pve shows I have:Code
5.4.103-1 is currently active, pending a reboot to switch to 5.4.106-1
Slightly off topic question, but does the extras functionality remove the older proxmox kernels?
(or is this a manual task?)
I've now got 10 proxmox/pve kernels and 10 recovery images showing in the dropdown/boot screen.
(also had a random crash yesterday, with maxed out IO-WAIT - not sure what occurred/where - but it rebooted to the latest (yesterday) kernel and then downloaded another one today
Thanks (and I see similar from that poster and another now) - I did a search for these packages before I posted and didn't get a response (mind you I got an error when I submitted the post too, so...)
I shall leave it do it's thing.
I have a salient tale of "not paying attention" before on a previous OMV install when I installed ntpd and it broke *all the things*. Stuff lives in Docker now!
...will show what is installed from backports, which (in my case) is:
I've noticed in the nightly update checks that:
So, given that sometimes they need a specific poke, I tried
And this horrifying message was returned....Code
Oddly enough, I don't want half the valid OS (and OMV!) being ripped out, so I hit CTRL-C
I noticed e2fsprogs was saying it would be removed, so checked that one:Code
I did wonder if a dodgy repo was at fault, but from what I can see with apt-cache showpkg on either package, both are coming direct out of the core debian repos (I DO have an HP repo available for my microserver's agentless management service).
A double check:Code
...shows they are the main repos - albeit backports
I have proxmox installed via omx-extras - and uname-r shows
How do I get out of this one? (or is it a waiting game?)
Your local network will probably work as your router will be intelligent and "hairpin" the external address back to the WAN interface and process accordingly.
My OMV is on port 81 (similar reason to you) but I'm using Traefik to handle the proxying (with SSL) and it works.
External IP:443->Router->Port Forward to Traefik Docker IP:443->SNI->OMV:81
Here is my Wireguard config. As mentioned, it runs in BRIDGE mode (set via Portainer). I access it via Traefik (i.e. my router forwards the WG port to my host, and Traefik handles it from there). I use a common DNS name internally and externally, BOTH resolve (internally and externally) to my public IP, and my router handles the hairpin.Code
My WireGuard config is:Code
I have set my CLIENT (Android) AllowedIPs to be:
And my CLIENT (Android) WireGuard DNS is my LAN DNS IP.
That way Wireguard is (AllowedIP) to see my DNS, the LAN (when DNS resolves internally) and "the internet".
If you remove 0.0.0.0/0 you won't get internet over WG, which is fine if you don't set "Block Connections Without VPN" in Android - i.e. you'll split tunnel.
If you're not clear on Traefik, you might want to read my main guide on OMV/Portainer/Traefik which is here : https://site.gothtech.co.uk/ar…ainer-traefik-letsencrypt
And then read my WireGuard add-on which is here: https://site.gothtech.co.uk/ar…encrypt/wireguard-traefik
I've shoved an OMV specific blog about this over here : https://sites.google.com/view/…authuser=0#h.o4dayltofvee
Portainer 2 also now fixes the issue I reported with not excluding the "shim" IP from the macvlan range - which is nice
I still use it, albeit not often, although I've also got Jellyfin running so I can see two DLNA options - not retired minidlna yet
I did install it at roughly the same time
You need to point a jail at your emby log or reverse proxy log
I use Traefik and just point it at the access.log
This should work if you're using Traefik (and could be adapted for other logs relatively easily)
Your problem is this:
s6-applyuidgid: fatal: unable to exec php7: Permission denied
OMV mounts the data volume with noexec
Docker is probably using a volume off that partition, and OMV won't let anything execute.
You need to remove noexec from the fstab entry (ideally via the OMV config.xml)
Or you can see how to on my blog post here:
Your settings will survive, and as long as you've setup volumes correctly, your data should persist.
If your container takes a while to init, this can get annoying if you miss something (dlandon/zoneminder takes about 10 mins to start...) but overall it's good. I really prefer the Portainer UI to the old docker one - and I've only had problems when I installed Traefik, setup a DNS entry for portainer to use Traefik (with SSL) and then took Traefik down (I forgot what I did, so that was fun).