Problem with pveca "missing ip address for node"

[Robstarusa]

Hello!

A couple of Q's. I have 2 servers (not so powerful, but they work ok) to try to create a cluster.

The name of these are "vm2" and "vm"

Configuration is as follows:

vm2: 5 * 400G raid10, 1 spare, 300G boot drive, amd athlon 5050e
vm1: 4 * 500g raid10, 1 80g boot drive, amd athlon 4850e

vm2 has 5 vm's running on it fairly well.
vm1 I just built tonight.

After installation, the arrays are turned on. I run pvcreate on them and then vgextend pve and then lvextend /data to 179200 extents (700.0GB). I do this for both machine

so far so good.

on vm1, I have configured 5 nics. Not all of them are currently hooked up. I have created

vmbr0 -> eth0
.
.
.
vmbr4 -> vmbr4

Currently vmbr4 is the active one with an IP:

vm2:~# ifconfig vmbr4
vmbr4     Link encap:Ethernet  HWaddr 00:19:66:A3:11:FC  
          inet addr:192.168.75.29  Bcast:192.168.75.255  Mask:255.255.255.0
          inet6 addr: fe80::219:66ff:fea3:11fc/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:47156272 errors:0 dropped:0 overruns:0 frame:0
          TX packets:36863619 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:37950266082 (35.3 GiB)  TX bytes:43229208201 (40.2 GiB)

I tried to start the cluster by following the faq (http://pve.proxmox.com/wiki/Proxmox_VE_Cluster)

I ran pveca -c and it created some ssh keys, I think, but now if I do pveca -l I get:

vm2:~# pveca -l
Use of uninitialized value in concatenation (.) or string at /usr/bin/pveca line 128.
local node '' not part of cluster

Runnig pveca -c again gives me:

vm2:~# pveca -c
missing ip address for node '1'

What can I do to fix this? I'm fairly good with linux so I don't mind doing anyting at the cli.

Thanks,

Rob

 

[fromport]

vm2:~# pveca -c
missing ip address for node '1'

Is there a definition in /etc/hosts for "vm2" ?

It looks (to me) it cannot resolve it's own name/host address.

So i the contents of /etc/hostname should be defined somewhere in /etc/hosts

Just my guess (new to pve)

Fromport

 

[Robstarusa]

Is there a definition in /etc/hosts for "vm2" ?

It looks (to me) it cannot resolve it's own name/host address.

So i the contents of /etc/hostname should be defined somewhere in /etc/hosts

Just my guess (new to pve)

Fromport

Thanks for the lightning fast reply. my nodename "vm2" is in /etc/hosts

 

[fromport]

Thanks for the lightning fast reply. my nodename "vm2" is in /etc/hosts

Okay, than that's not it.Sorry have to wait till the people on the other side of the pond wake up ;-)
Only have a single machine running yet (and have my own obstacles ;-) )
Am not able to "help" you. sorry.

Fromport

 

[dietmar]

You need to assign an IP to vmbr1 (or eth0). We use that IP as primary IP.

 

[dietmar]

vm2: 5 * 400G raid10, 1 spare, 300G boot drive, amd athlon 5050e
vm1: 4 * 500g raid10, 1 80g boot drive, amd athlon 4850e

IMO it makes absolutely no sense to use a separate root partition. Instead you should install everything on RAID10.

At lease use RAID for the root partition as well.

- Dietmar

 

[Robstarusa]

IMO it makes absolutely no sense to use a separate root partition. Instead you should install everything on RAID10.

At lease use RAID for the root partition as well.

- Dietmar

I did end up doing thiis....I will post tonight (I'm GMT -6) after work so perhaps I can help other people.