how to restore Proxmox configuration files

[bravo0916]

Hello team,

I only have one Proxmox host and OS is being installed on one SSD disk. There is no RAID. However, all datas have redundancy like ZFS.
I think the data will be OK, but I am concerned with OS. If I setup new Proxmox again in the same SSD, then I need to setup network (virtual network, bond network, etc), hostname, storage configuration, etc. I am not sure if there are something like configuration files and I just download them and upload into new Proxmox. After that, new Proxmox will recognized these configurations that old Proxmox had them. And I don't need to configure again. For FreeNAS, it is possible and it is very easy.
Sorry, I understand that Proxmox and Freenas system architecher is different.

Also, if I add one more Proxmox, then both Proxmox are supposed to be replicated all configuration data like network, storage, hostname, etc right?
Let's suppose that there are NodeA and NodeB. If NodeA is down and I try to setup NodeA from scratch and join into NodeB cluster, then NodeA has the same configuration files from NodeB. In this case, I don't need to setup everything from scratch. Is this correct?

Regards,

 

[spirit]

backup /var/lib/pve-cluster/config.db

- install a new proxmox with same hostname / ip address
- boot
- systemctl stop pve-cluster
- copy back /var/lib/pve-cluster/config.db
-reboot

done


(if you are in a cluster, copy back also /etc/corosync/* )

 

[bravo0916]

Hi spirit,

Thanks for your quick reply and I would like to thank you for your support. I will try this later.

Regards,

 

[esi_y]

For more details and ideas, you can also have a look here:

https://forum.proxmox.com/threads/backup-cluster-config-pmxcfs-etc-pve.154569/

https://forum.proxmox.com/threads/r...ffline-extract-configurations-etc-pve.155374/

 

[esi_y]

I only have one Proxmox host and OS is being installed on one SSD disk. There is no RAID. However, all datas have redundancy like ZFS.
I think the data will be OK, but I am concerned with OS.

The data will not be okay without backups, either.

See e.g.:

https://forum.proxmox.com/threads/z...o-error-on-import-attempt.156075/#post-712516

These are actually quite common. It helps not using ZFS for root, at the least.

Also, if I add one more Proxmox, then both Proxmox are supposed to be replicated all configuration data like network, storage, hostname, etc right?

Correct, but it's quite a brittle setup with just 2 nodes and no special configuration or a Q device, see also:

https://pve.proxmox.com/wiki/Cluster_Manager

Let's suppose that there are NodeA and NodeB. If NodeA is down and I try to setup NodeA from scratch and join into NodeB cluster, then NodeA has the same configuration files from NodeB.

This does not work as easy as stated above due to the 2-only setup.

In this case, I don't need to setup everything from scratch. Is this correct?

But consider this is not a backup by any means, if your config gets ruined, it gets ruined on both (all) nodes instantly.

 

[bravo0916]

Hi esi_y,

Thank you so much for your good advice! Yes, I still need to take backup for guest OS, etc.. I am using Veeam backup for this.
Now, let me ask you about ZFS (raid1) please. You mean, root disk should not used raid1? I can see document that ZFS raid1 is one of the options for root disk. https://pve.proxmox.com/wiki/ZFS_on_Linux
What do you suggest for root disk?

Regards,

 

[esi_y]

Thank you so much for your good advice! Yes, I still need to take backup for guest OS, etc.. I am using Veeam backup for this.

Not being a party pooper, but currently, there's this active thread:
https://forum.proxmox.com/threads/veeam-silent-data-corruption.155212/

Now, let me ask you about ZFS (raid1) please. You mean, root disk should not used raid1? I can see document that ZFS raid1 is one of the options for root disk. https://pve.proxmox.com/wiki/ZFS_on_Linux

Proxmox basically offers only ZFS (from GUI) for RAIDs, not counting BTRFS. There's other options, especially if you do your own install, i.e. on top of Debian [1].

I believe the big question is, for e.g. homelab user, what is the point of RAID, if it's not high availability (and then cluster what is the purpose of cluster), then not using RAID at all is something I would advise. Whatever I advise, there will be 5 people on this forum against it.

What do you suggest for root disk?

I really do not see point of ZFS on root - that's my entirely separate point (as it causes issues which were real in the linked thread). There's nothing much valuable on root that you cannot just reinstall from ISO except the configs (you had asked about above). So you can do Debian install and have whatever Debian offers, or you can e.g. do Proxmox VE ISO install, choose LVM w/ext4 (or XFS) for root and leave it at that.

NB There's even the quirk that if you choose ZFS install for root on UEFI and not SB, then it gives you systemd-boot instead of GRUB bootloader even..

I want to point out, even if you really want ZFS for the rest, you can still install as mentioned above, then add ZFS pool for the guests only (even RAID). The stock installer [2] has even "advanced" configuration options for disk where you can specify e.g. hdsize and space beyond that will be unpartitioned (so you can use it for the ZFS that you create post-installation all via GUI).

I do not like ZFS, but those discussions already happened here (easy to search threads) and I am apparently in minority and basically it goes against the choice Proxmox once made to streamline their support. All I wanted to let you know was - ZFS can destroy your data, too.

[1] https://pve.proxmox.com/wiki/Install_Proxmox_VE_on_Debian_12_Bookworm
[2] https://pve.proxmox.com/wiki/Installation

 

[bravo0916]

Hi sprit,

I installed new Proxmox and version is 8.2.2. I downloaded it from Proxmox site. And I copied back config.db, but I can't connect to https://xxxx:8006 (Proxmox dashboard). I can see that "/etc/pve/local/pve-ssl.key: failed to load local privte key..." when I executed journalctl -xe command. The ping result was OK. The Proxmox can be reachable to gateway and I can ping from my PC to Proxmox. But I got the time out when I accessed to Proxmox dashboard. Also, I can see that vmbr0: port 1 (eno1) entered blocking state. My previous Proxmox version was 8.2.4 that I updated. However, current ISO is 8.2.2. I am not sure if this is the root cause or not.

Regards,

 

[esi_y]

Hi sprit,

I installed new Proxmox and version is 8.2.2. I downloaded it from Proxmox site. And I copied back config.db, but I can't connect to https://xxxx:8006 (Proxmox dashboard). I can see that "/etc/pve/local/pve-ssl.key: failed to load local privte key..." when I executed journalctl -xe command. The ping result was OK. The Proxmox can be reachable to gateway and I can ping from my PC to Proxmox. But I got the time out when I accessed to Proxmox dashboard. Also, I can see that vmbr0: port 1 (eno1) entered blocking state. My previous Proxmox version was 8.2.4 that I updated. However, current ISO is 8.2.2. I am not sure if this is the root cause or not.

Regards,

Is it THIS issue?

https://forum.proxmox.com/threads/f...-key-failed-to-load-local-private-key.155771/

 

[bravo0916]

Hi esi_y,

I tried to reinstall again, but I still can't connect to 8006. I don't see pve-ssl.key error in this time. Mmmm. I still can ping to Proxmox.
I have not copied back previous config.db file yet.

Regards,

 

[esi_y]

Hi esi_y,

I tried to reinstall again, but I still can't connect to 8006. I don't see pve-ssl.key error in this time. Mmmm. I still can ping to Proxmox.
I have not copied back previous config.db file yet.

Regards,

So what is in the journalctl -e -u pveproxy -u pve-cluster ?

 

[bravo0916]

I can access to Proxmox via SSH connection.

 

[bravo0916]

Here is the result.

Oct 20 17:11:32 jptyovm02 systemd[1]: Starting pve-cluster.service - The Proxmox VE cluster filesystem...
Oct 20 17:11:32 jptyovm02 pmxcfs[1143]: [main] notice: resolved node name 'jptyovm02' to '192.168.30.88' for default node IP address
Oct 20 17:11:32 jptyovm02 pmxcfs[1143]: [main] notice: resolved node name 'jptyovm02' to '192.168.30.88' for default node IP address
Oct 20 17:11:33 jptyovm02 systemd[1]: Started pve-cluster.service - The Proxmox VE cluster filesystem.
Oct 20 17:11:34 jptyovm02 systemd[1]: Starting pveproxy.service - PVE API Proxy Server...
Oct 20 17:11:35 jptyovm02 pveproxy[1284]: starting server
Oct 20 17:11:35 jptyovm02 pveproxy[1284]: starting 3 worker(s)
Oct 20 17:11:35 jptyovm02 pveproxy[1284]: worker 1285 started
Oct 20 17:11:35 jptyovm02 pveproxy[1284]: worker 1286 started
Oct 20 17:11:35 jptyovm02 pveproxy[1284]: worker 1287 started
Oct 20 17:11:35 jptyovm02 systemd[1]: Started pveproxy.service - PVE API Proxy Server.

 

[esi_y]

Can you double-check:

ip a

And see what curl says (on the node itself):

curl -k https://localhost:8006

Also, are you positive you do not have e.g. IP conflict? Because the "beauty" of PVE is in the "static" silliness...

 

[bravo0916]

Hi esi_y,

There is no IP conflic for sure. Here is the result.

root@jptyovm02:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host noprefixroute
valid_lft forever preferred_lft forever
2: enp1s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
link/ether 70:85:c2:af:94:a1 brd ff:ff:ff:ff:ff:ff
3: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master vmbr0 state UP group default qlen 1000
link/ether 70:85:c2:af:94:9f brd ff:ff:ff:ff:ff:ff
altname enp0s31f6
4: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 70:85:c2:af:94:9f brd ff:ff:ff:ff:ff:ff
inet 192.168.30.88/24 scope global vmbr0
valid_lft forever preferred_lft forever
inet6 fe80::7285:c2ff:feaf:949f/64 scope link
valid_lft forever preferred_lft forever


root@jptyovm02:~# curl -k https://localhost:8006
<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1, maximum-scale=1, user-scalable=no">
<title>jptyovm02 - Proxmox Virtual Environment</title>
<link rel="icon" sizes="128x128" href="/pve2/images/logo-128.png" />
<link rel="apple-touch-icon" sizes="128x128" href="/pve2/images/logo-128.png" />
<link rel="stylesheet" type="text/css" href="/pve2/ext6/theme-crisp/resources/theme-crisp-all.css?ver=7.0.0" />
<link rel="stylesheet" type="text/css" href="/pve2/ext6/crisp/resources/charts-all.css?ver=7.0.0" />
<link rel="stylesheet" type="text/css" href="/pve2/fa/css/font-awesome.css" />
<link rel="stylesheet" type="text/css" href="/pve2/font-logos/css/font-logos.css" />
<link rel="stylesheet" type="text/css" href="/pve2/css/ext6-pve.css?ver=8.2.2" />
<link rel="stylesheet" type="text/css" href="/pwt/css/ext6-pmx.css?ver=v4.2.1-t1713893106" />
<link rel="stylesheet" type="text/css" media="(prefers-color-scheme: dark)" href="/pwt/themes/theme-proxmox-dark.css?ver=v4.2.1-t1713893106" />

<script type='text/javascript'>function gettext(buf) { return buf; }</script>

<script type="text/javascript" src="/pve2/ext6/ext-all.js?ver=7.0.0"></script>
<script type="text/javascript" src="/pve2/ext6/charts.js?ver=7.0.0"></script>

<script type="text/javascript" src="/pve2/js/u2f-api.js"></script>
<script type="text/javascript" src="/qrcode.min.js"></script>
<script type="text/javascript">
Proxmox = {
Setup: { auth_cookie_name: 'PVEAuthCookie' },
defaultLang: 'en',
NodeName: 'jptyovm02',
UserName: '',
CSRFPreventionToken: 'null'
};
</script>
<script type="text/javascript" src="/proxmoxlib.js?ver=v4.2.1-t1713893106"></script>
<script type="text/javascript" src="/pve2/js/pvemanagerlib.js?ver=8.2.2"></script>
<script type="text/javascript" src="/pve2/ext6/locale/locale-en.js?ver=7.0.0"></script>

<script type="text/javascript">
if (typeof(PVE) === 'undefined') PVE = {};
Ext.History.fieldid = 'x-history-field';
Ext.onReady(function() { Ext.create('PVE.StdWorkspace');});
</script>

</head>
<body>
<!-- Fields required for history management -->
<form id="history-form" class="x-hidden">
<input type="hidden" id="x-history-field"/>
</form>
</body>
</html>

 

[esi_y]

I actually do not see anything wrong, your interface has the IP, the pveproxy serves the content. If you have another Linux you can try to access from, you might want curl from there?

 

[bravo0916]

I don't have any other Linux machines except guestOS. Currently, I can not launch all of them because they are running on Proxmox.
Now, I found something that I don't see access.log under /var/log/pveproxy directory. I tried to install Proxmox 8.2.2 ISO to the other disk, but the situation is the same. Is this something bug???


systemctl status pveproxy.service pvedaemon.service
● pveproxy.service - PVE API Proxy Server
Loaded: loaded (/lib/systemd/system/pveproxy.service; enabled; preset: enabled)
Active: active (running) since Sun 2024-10-20 19:25:37 JST; 40min ago
Process: 1093 ExecStartPre=/usr/bin/pvecm updatecerts --silent (code=exited, status=0/SUCCESS)
Process: 1111 ExecStart=/usr/bin/pveproxy start (code=exited, status=0/SUCCESS)
Main PID: 1112 (pveproxy)
Tasks: 4 (limit: 154089)
Memory: 170.1M
CPU: 2.782s
CGroup: /system.slice/pveproxy.service
tq1112 pveproxy
tq1113 "pveproxy worker"
tq1114 "pveproxy worker"
mq1115 "pveproxy worker"

Oct 20 19:25:36 jptyovm02 pvecm[1097]: |. ..oo+=. . |
Oct 20 19:25:36 jptyovm02 pvecm[1097]: | oEo..+o++o |
Oct 20 19:25:36 jptyovm02 pvecm[1097]: +----[SHA256]-----+
Oct 20 19:25:37 jptyovm02 pvecm[1094]: got inotify poll request in wrong process - disabling inotify
Oct 20 19:25:37 jptyovm02 pveproxy[1112]: starting server
Oct 20 19:25:37 jptyovm02 pveproxy[1112]: starting 3 worker(s)
Oct 20 19:25:37 jptyovm02 pveproxy[1112]: worker 1113 started
Oct 20 19:25:37 jptyovm02 pveproxy[1112]: worker 1114 started
Oct 20 19:25:37 jptyovm02 pveproxy[1112]: worker 1115 started
Oct 20 19:25:37 jptyovm02 systemd[1]: Started pveproxy.service - PVE API Proxy Server.

● pvedaemon.service - PVE API Daemon
Loaded: loaded (/lib/systemd/system/pvedaemon.service; enabled; preset: enabled)
Active: active (running) since Sun 2024-10-20 19:45:27 JST; 20min ago
Process: 4346 ExecStart=/usr/bin/pvedaemon start (code=exited, status=0/SUCCESS)
Main PID: 4348 (pvedaemon)
Tasks: 4 (limit: 154089)
Memory: 145.3M
CPU: 696ms
CGroup: /system.slice/pvedaemon.service
tq4348 pvedaemon
tq4349 "pvedaemon worker"
tq4350 "pvedaemon worker"
mq4351 "pvedaemon worker"

Oct 20 19:45:27 jptyovm02 systemd[1]: Starting pvedaemon.service - PVE API Daemon...
Oct 20 19:45:27 jptyovm02 pvedaemon[4348]: starting server
Oct 20 19:45:27 jptyovm02 pvedaemon[4348]: starting 3 worker(s)
Oct 20 19:45:27 jptyovm02 pvedaemon[4348]: worker 4349 started
Oct 20 19:45:27 jptyovm02 pvedaemon[4348]: worker 4350 started
Oct 20 19:45:27 jptyovm02 pvedaemon[4348]: worker 4351 started
Oct 20 19:45:27 jptyovm02 systemd[1]: Started pvedaemon.service - PVE API Daemon.

root@jptyovm02:~# cd /var/log/pveproxy/
root@jptyovm02:/var/log/pveproxy# ls
root@jptyovm02:/var/log/pveproxy# ls
root@jptyovm02:/var/log/pveproxy#

 

[bravo0916]

Hello.

I tried to recreate pveproxy directory and restart pveproxy.service. Now, I can see access.log file, but this file is empty even if I tried to access to https://xxxx:8006.

 

[esi_y]

Hello.

I tried to recreate pveproxy directory and restart pveproxy.service. Now, I can see access.log file, but this file is empty even if I tried to access to https://xxxx:8006.

I was kind of trying to say, you are failing to access from the outside, the pveproxy appears to be running and doing what you want, so you need to troubleshoot the connection.

I am sure this could be done e.g. with some tooling from Windows/Mac, but I would just e.g. install VirtualBox or UTM, put in stock Debian server (no GUI) install and start checking from there. Some routers also have certain toolings. Otherwise I would need to be suggesting things like clear browser cache, etc. I think your issue is not on the install, but in the network, from what you described above.

 

[bravo0916]

Hello esi_y,

I am just doing the test and I want to confirm if config.db can work or not when I setup new Proxmox system. Now, I switched to production SSD disk which has good Proxmox. I can see that I can access to Proxmox without any issues. Also, I tried to install TrueNAS and I can access to Web Gui.
I used the same IP and gateway, etc. Thus, I don't think that the issue is network side. I double checked with my router setting especially port 8006, and any other necessary ports for Proxmox. Everthing looks good. Is there any older version image file of Proxmox? I want to try.
The install USB disk and target SSD should be fine since Trunas has been installed successfully and access works.