Request for help cleaning up my config.db

[dommel]

Hi there!
I am somewhat of a fresh home user of proxmox.
While adding a second host to my cluster, my first hostname had to change. After following the KB on this (edit /etc/hosts /etc/hostname and /etc/postfix/main.cf and the coroconf) a reboot I noticed this host with the wrong (old) name in the UI. (I did follow the same procedure on another host, and that rename went just fine...) So I tried figuring out how to fix that as well. I noticed that all my VMs were still in the folder under /etc/pve/nodes/ so I moved the qemu-server folder from the old server named folder to the new named folder.
After another reboot my server would not come back online.

I traced it to a duplicate entry in the config.db. But I am rly not db savy and confident enough to edit this...

journalctl -b -u pve-cluster gives me (among other lines):

systemd[1]: Starting The Proxmox VE cluster filesystem...
pmxcfs[1295]: [database] crit: found entry with duplicate name 'qemu-server' - A:(inode = 0x000000000030E094, parent = 0x000000000030E093, v./mtime = 0x30E094/0x1671282833) vs. B:(inode = 0x0000000000>
pmxcfs[1295]: [database] crit: found entry with duplicate name 'qemu-server' - A:(inode = 0x000000000030E094, parent = 0x000000000030E093, v./mtime = 0x30E094/0x1671282833) vs. B:(inode = 0x0000000000>
pmxcfs[1295]: [database] crit: DB load failed
pmxcfs[1295]: [main] crit: memdb_open failed - unable to open database '/var/lib/pve-cluster/config.db'
pmxcfs[1295]: [main] notice: exit proxmox configuration filesystem (-1)
pmxcfs[1295]: [database] crit: DB load failed
pmxcfs[1295]: [main] crit: memdb_open failed - unable to open database '/var/lib/pve-cluster/config.db'
pmxcfs[1295]: [main] notice: exit proxmox configuration filesystem (-1)

pmcxfs -d

[database] crit: found entry with duplicate name 'qemu-server' - A:(inode = 0x000000000030E094, parent = 0x000000000030E093, v./mtime = 0x30E094/0x1671282833) vs. B:(inode = 0x000000000030E64B, parent = 0x000000000030E093, v./mtime = 0x30E64B/0x1671283925) (database.c:430:bdb_backend_load_index)
[database] crit: DB load failed (database.c:470:bdb_backend_load_index)
[main] crit: memdb_open failed - unable to open database '/var/lib/pve-cluster/config.db' (pmxcfs.c:888:main)
[main] notice: exit proxmox configuration filesystem (-1) (pmxcfs.c:1099:main)

Could someone help me figure out how to remove the duplicate entry in there?

 

[dommel]

I did not realize and feel stupid for it, but quite some important stuff is not backed up on the vms on this host. I did not think renaming the host could actually break it this bad I am a little stressed now... Is there a way to extract the vm's from this host? I have other hosts I can run this on. But it being down is not something I can live with long.

Btw I noticed that

uname -a

outputs the old hostname still.

 

[dommel]

I tried to follow along in this topic: https://forum.proxmox.com/threads/pve-cluster-fails-to-start.82861/

# sqlite3 /var/lib/pve-cluster/config.db 'PRAGMA integrity_check'
ok
# sqlite3 /var/lib/pve-cluster/config.db .schema
CREATE TABLE tree (  inode INTEGER PRIMARY KEY NOT NULL,  parent INTEGER NOT NULL CHECK(typeof(parent)=='integer'),  version INTEGER NOT NULL CHECK(typeof(version)=='integer'),  writer INTEGER NOT NULL CHECK(typeof(writer)=='integer'),  mtime INTEGER NOT NULL CHECK(typeof(mtime)=='integer'),  type INTEGER NOT NULL CHECK(typeof(type)=='integer'),  name TEXT NOT NULL,  data BLOB);
# sqlite3 /var/lib/pve-cluster/config.db 'SELECT inode,mtime,name FROM tree WHERE parent = 0'
0|1671283999|__version__
2|1666536744|storage.cfg
6|1666536744|datacenter.cfg
8|1666536786|virtual-guest
9|1666536787|priv
11|1666536787|nodes
24|1666536787|pve-www.key
30|1666536787|pve-root-ca.pem
49|1666536787|ha
51|1666536787|sdn
112995|1666704146|replication.cfg
112999|1666704146|vzdump.cron
2232248|1669902564|user.cfg
3098910|1671202556|authkey.pub.old
3098913|1671202556|authkey.pub
3100476|1671283007|corosync.conf
# sqlite3 /var/lib/pve-cluster/config.db 'SELECT inode,mtime,name FROM tree WHERE parent = 347747 or inode = 347747'
#

 

[dommel]

sqlite3 /var/lib/pve-cluster/config.db 'SELECT inode,mtime,name FROM tree WHERE parent = 0x000000000030E093 or inode = 0x000000000030E094'
13|1671283732|lxc
15|1671283732|openvz
16|1671283732|priv
26|1671283732|pve-ssl.key
35|1671283732|pve-ssl.pem
3203220|1671282833|qemu-server
3204683|1671283925|qemu-server
3204785|1671283996|lrm_status

Schould I delete one of these? And how do I know which one is the old one?

 

[dommel]

# sqlite3 /var/lib/pve-cluster/config.db 'SELECT inode,parent,mtime,type,name FROM tree WHERE inode = 3203219'
3203219|11|1671282833|4|new-host-name
# sqlite3 /var/lib/pve-cluster/config.db 'SELECT inode,parent,mtime,type,name FROM tree WHERE inode = 3203220'
3203220|3203219|1671282833|4|qemu-server

I am still unsure what I am looking at. I did create a backup, so I think I'll just delete one and see what happens.

 

[dommel]

sqlite3 /var/lib/pve-cluster/config.db 'DELETE FROM tree WHERE inode = 3203220'

This did do something. At least it is starting now. Still had it show with the old and the new hostname in UI. But I guess I'll take it...

 

[forted]

Thank you so much, this save me today.
If you still have a problem with a duplicate hostname in UI, I can help.

 

[lknhd]

Dude, i just made this account just to say thank you. You saved my day. Thank you so much.

 

[Lqdtr]

Same problem. dommel, thank you very much!!!!