pve configuration filesystem not mounted

[Posi]

I killed my cluster and need to get my VM's back on-line.
How do you delete the cluster so you can use the system again and run is solo?

 

[dietmar]

http://pve.proxmox.com/wiki/Cluster_Manager#pvecm_separate_node_without_reinstall

 

[Posi]

I have been looking at that thread.

The instructions from the link

systemctl stop pve-cluster # not needed as the cluster will not start
systemctl stop corosync # same. not started, wish it was

# next steep run pmxcfs -l to run the pve file system. See Error

root@prox1:~# pmxcfs -l
[main] notice: forcing local mode (althought corosync.conf exists)
fuse: mountpoint is not empty
fuse: if you are sure this is safe, use the 'nonempty' mount option
[main] crit: fuse_mount error: File exists
[main] notice: exit proxmox configuration filesystem (-1)
root@prox1:~#

# need to nonempty. So let's look at the fuse.conf file.

# /etc/fuse.conf - Configuration file for Filesystem in Userspace (FUSE)

# Set the maximum number of FUSE mounts allowed to non-root users.
# The default is 1000.
#mount_max = 1000

# Allow non-root users to specify the allow_other or allow_root mount options.
#user_allow_other

root@prox1:~# journalctl -xn
-- Logs begin at Sun 2016-11-13 06:36:17 EST, end at Sun 2016-11-13 06:58:40 EST. --
Nov 13 06:58:38 prox1 pve-ha-crm[2105]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:38 prox1 pve-ha-lrm[2114]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:38 prox1 pve-ha-lrm[2114]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:38 prox1 pve-ha-lrm[2114]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused
Nov 13 06:58:40 prox1 pvestatd[1645]: ipcc_send_rec failed: Connection refused

##Looks like the cluster is trying to connect with the other nodes with the bad key.


root@prox1:~# pmxcfs -l -d
[database] debug: name __version__ (inode = 0000000000000000, parent = 0000000000000000) (database.c:373:bdb_backend_load_index)
[database] debug: name corosync.conf (inode = 0000000000000002, parent = 0000000000000000) (database.c:373:bdb_backend_load_index)
[main] debug: memdb open '/var/lib/pve-cluster/config.db' successful (version = 0000000000000003) (memdb.c:535:memdb_open)
[main] notice: forcing local mode (althought corosync.conf exists) (pmxcfs.c:875:main)
fuse: mountpoint is not empty
fuse: if you are sure this is safe, use the 'nonempty' mount option
[main] crit: fuse_mount error: File exists (pmxcfs.c:906:main)
[main] notice: exit proxmox configuration filesystem (-1) (pmxcfs.c:1029:main)

# I can't get the file system to start ;(

 

[dietmar]

root@prox1:~# pmxcfs -l
[main] notice: forcing local mode (althought corosync.conf exists)
fuse: mountpoint is not empty

Seems /etc/pve is not mounted, but there is simply some data inside "/etc/pve/"? Please move that data to another place first.

 

[Posi]

Because CMAP can't run all the files are locked and I can't delete them.
How can I delete them?

 

[franko5]

Seems /etc/pve is not mounted, but there is simply some data inside "/etc/pve/"? Please move that data to another place first.

Hello, I know it is old topic, and that fuse is used to load the data into directory. I see the explanation on manual here

Still, can you please help here, I have a strange situation and had to recover quite some files, when I recovered from backup /etc/pve, pve-cluster.service will not start.
When I remove the content of /etc/pve - as you suggested, pve-cluster is starting, but after start /etc/pve is still empty.
Also, the gui is working as well as VM's.
Now, what is the proper way to restore /etc/pve?
The database in /var/lib/pve-cluster/config.db looks good, there are data about used certificates etc.

 

[franko5]

When I remove the content of /etc/pve - as you suggested, pve-cluster is starting, but after start /etc/pve is still empty.

After a while, the content is again within /etc/pve, as expected, seems that it is automatically populated after a while. Looks fine again.