new node wont join cluster

[suprstar]

I have 3 servers already set up in a cluster:

root@proxmox3:~# pvecm n
Node  Sts   Inc   Joined               Name
   1   M    124   2013-01-05 16:45:57  proxmox2
   2   M    124   2013-01-05 16:45:57  proxmox3
   3   M    272   2013-02-08 12:04:01  proxmox4

I'm trying to add a 4th. When I get an error and have to use -force, probably due to all the tinkering I've done trying to get this working... With -force, I get another error "unable to copy ssh ID":

root@proxmox5:~# pvecm add proxmox3
authentication key already exists
root@proxmox5:~# pvecm add proxmox3 -force
unable to copy ssh ID
root@proxmox5:~#

In my searching, I've found it suggested I try 'ssh-copy-id' on a command line to see why it's failing, here it is:

root@proxmox5:~# ssh-copy-id proxmox3
cat: write error: File too large
root@proxmox5:~#

I have plenty of space, enough for ssh keys at least:

root@proxmox5:~# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/system-root
                      3.0G  481M  2.6G  16% /
tmpfs                 3.9G     0  3.9G   0% /lib/init/rw
udev                  3.9G  288K  3.9G   1% /dev
tmpfs                 3.9G   35M  3.9G   1% /dev/shm
/dev/md0              289M   93M  182M  34% /boot
/dev/mapper/system-home
                      508M   26M  482M   6% /home
/dev/mapper/system-opt
                      508M   26M  482M   6% /opt
/dev/mapper/system-tmp
                     1014M   33M  982M   4% /tmp
/dev/mapper/system-usr
                     1014M  459M  556M  46% /usr
/dev/mapper/system-var
                     1014M  407M  608M  41% /var
/dev/mapper/system-images
                      300G   33M  300G   1% /srv/vm/images
/dev/mapper/system-isos
                       50G   33M   50G   1% /srv/vm/isos
/dev/mapper/system-var_log
                      508M   36M  472M   8% /var/log
/dev/fuse              30M   16K   30M   1% /etc/pve
root@proxmox5:~#

I'm at a loss how to troubleshoot this further, any ideas?

 

[suprstar]

In case anyone searches and finds this later - Turns out there was a few hundred lines in .ssh/authorized_keys, my co uses key-based access to everything. Too many and it makes that little script bomb, who knew...... I took out all the keys except the ones for the cluster proxmox servers (and my own), and it worked just fine after that.

 

[mir]

You have been strucked by an old bug. At some point an upgrade caused pollution of the known_hosts file.