Web UI problem after OMV upgrade 6to7

System:

Proxmox virtualized, Intel x86

HBA passtrough with ZFS mirror on disks

OMV6 latest (now 7) version with Proxmox kernel

After doing the upgrade to OMV7 I noticed that after server startup I cant open web UI. I need to wait at least 2-3 minutes and than the web UI will be operational.

The problem only shows after a start/reboot of the system.

Got the alert emails with messages:

Host: nas
Date: Wed, 13 Mar 2024 01:51:04

Service: nginx
Event: Does not exist
Description: process is not running
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:51:34

Service: nginx
Event: Execution failed
Description: failed to start (exit status -1) -- Program '/bin/systemctl start nginx' timed out after 30 s
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:52:27

Service: nginx
Event: Execution succeeded
Description: started
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:52:58

Service: nginx
Event: Exists
Description: process is running with pid 1293

What steps can I take to remedy this problem?

PS: After update I've also noticed that NUT service was flooding the logs with "UPS unavailable" messages.

The thing is that I did not use NUT for a year (removed it from omvextras after I connected the UPS to the host system). So somehow it got installed in the process of upgrading OMV 6 to 7. I've removed it manually, and now it's OK.

Thanks and kind regards.

One more thing I noticed on my other server; when I install a plugin it does not get the green "installed" button.

iperf3 for example:

This is for the second server, plugin problem

- I did try with the cache cleared before, even used another browser - no good

- Ran the script (no errors), cleared the cache again - no good

- sudo omv-salt deploy run phpfpm (no errors):

Succeeded: 16 (changed=10)

Failed: 0

- sudo omv-salt deploy run nginx (no errors):

Succeeded: 18 (changed=11)

Failed: 0

sudo systemctl restart nginx (no errors)

sudo systemctl restart php8.2-fpm.service (no errors)

Cleared the cache and cookies again, no good.

I then tried to reinstall the iperf3 plugin again to see what will happen, this is the output:

EDIT:

So now I ran the "fix6to7upgrade" script again and refreshed - now its good!

It produced the same output as before, but after a refresh in browser everything looks fine.

Thanks macom !

Zitat von vperas

After doing the upgrade to OMV7 I noticed that after server startup I cant open web UI. I need to wait at least 2-3 minutes and than the web UI will be operational.

The problem only shows after a start/reboot of the system.

Got the alert emails with messages:

Code

Host: nas
Date: Wed, 13 Mar 2024 01:51:04

Service: nginx
Event: Does not exist
Description: process is not running
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:51:34

Service: nginx
Event: Execution failed
Description: failed to start (exit status -1) -- Program '/bin/systemctl start nginx' timed out after 30 s
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:52:27

Service: nginx
Event: Execution succeeded
Description: started
===================================================
Host: nas
Date: Wed, 13 Mar 2024 01:52:58

Service: nginx
Event: Exists
Description: process is running with pid 1293

Alles anzeigen

What steps can I take to remedy this problem?

Alles anzeigen

I'm still chasing this, it's not a critical problem since nginx starts after a minute or two, but it's bugging me

Zitat von macom

try to find the reason with

systemd-analyze blame

systemd-analyze critical-chain

Output for systemd-analyze blame (redacted to ones above 1sec):

root@nas:~# systemd-analyze blame
2min 204ms systemd-networkd-wait-online.service
   30.930s apt-daily.service
   30.650s apt-daily-upgrade.service
    5.226s openmediavault-issue.service
    3.832s zfs-import-cache.service
    2.676s folder2ram_startup.service
    2.230s docker.service
    1.558s systemd-modules-load.service
    1.253s smartmontools.service
    1.007s systemd-udev-settle.service

And output for systemd-analyze critical-chain

root@nas:~# systemd-analyze critical-chain
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.

graphical.target @2min 7.728s
└─multi-user.target @2min 7.728s
  └─getty.target @2min 7.728s
    └─openmediavault-issue.service @2min 2.501s +5.226s
      └─network-online.target @2min 2.466s
        └─network.target @9.548s
          └─wpa_supplicant.service @9.520s +27ms
            └─dbus.service @9.409s +55ms
              └─basic.target @9.381s
                └─folder2ram_startup.service @6.704s +2.676s
                  └─local-fs.target @6.676s
                    └─run-docker-netns-e5e929c5e90f.mount @2min 4.981s
                      └─local-fs-pre.target @1.092s
                        └─systemd-tmpfiles-setup-dev.service @1.071s +20ms
                          └─systemd-sysusers.service @549ms +521ms
                            └─systemd-remount-fs.service @482ms +51ms
                              └─systemd-journald.socket @446ms
                                └─-.mount @408ms
                                  └─-.slice @408ms

So the problem is that the server waits for the network to be fully started for some reason, and that's causing the delay.

I'll try to find out the cause.

macom thanks!

EDIT:

User UltimateByte found out the same thing on his end, and offered a workaround: LINK

Zitat von UltimateByte

Yep, definitely looks like the exact same issue! Glat it worked out for you as well!

Did you change any network interface in the process?

If not, I need to update my guide, as it would rather be an issue related to migration from OMV6 to OMV7.

I use two interfaces in OMV; one for OMV itself and the other (disabled in OMV) for mcvlans under docker.

Under config I just added:

[Service]
ExecStart=
ExecStart=/usr/lib/systemd/systemd-networkd-wait-online --ignore=enp3s0

Thank you for the detailed explanation in your thread - led me in the right direction!