corosync: using 'ring1_addr' parameter needs a configured ring 1 interface!

[Thomas Jagoditsch]

hi !

im building a new test cluster with 5.1 to try all the new features.

im getting "corosync: using 'ring1_addr' parameter needs a configured ring 1 interface!" while adding the first node to the cluster regardless of using hostnames or ips.

what i did:

• installed 3 basically identical machines w. 2 NICs
• edited /etc/network/interfaces, /etc/hosts and ssh'ed each other so that all machines see & reach each other via the 2 interfaces by hostname "test<0-2>" (seperate cluster net) and "test<n>lan" (the lan/vm net and failover cluster net)
• created the cluster on node test0 with 2 rings as described in https://pve.proxmox.com/wiki/Cluster_Manager
• verified corosync.conf

root@test1:~# pvecm add test0 -ring0_addr test1 -ring1_addr test1lan
corosync: using 'ring1_addr' parameter needs a configured ring 1 interface!
unable to add node: command failed (ssh test0 -o BatchMode=yes pvecm addnode test1 --force 1 --ring0_addr test1 --ring1_addr test1lan)
root@test1:~# pvecm add 10.66.2.21 -ring0_addr 10.66.2.22 -ring1_addr 10.166.2.22
corosync: using 'ring1_addr' parameter needs a configured ring 1 interface!
unable to add node: command failed (ssh 10.66.254.21 -o BatchMode=yes pvecm addnode test1 --force 1 --ring0_addr 10.66.2.22 --ring1_addr 10.166.2.22)

cluster and corosync seem to be ok:

root@test0:~# pvecm status
Quorum information
------------------
Date:             Fri Nov 10 09:55:04 2017
Quorum provider:  corosync_votequorum
Nodes:            1
Node ID:          0x00000001
Ring ID:          1/16
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   1
Highest expected: 1
Total votes:      1
Quorum:           1 
Flags:            Quorate

Membership information
----------------------
    Nodeid      Votes Name
0x00000001          1 10.66.2.21 (local)

root@test0:~# corosync-cfgtool -s
Printing ring status.
Local node ID 1
RING ID 0
    id    = 10.66.2.21
    status    = ring 0 active with no faults
RING ID 1
    id    = 10.166.2.21
    status    = ring 1 active with no faults

i see nothing helpful or strange in the syslogs of either machine so im kinda stuck.
maybe somebody can give me a hint what im doing wrong here.

tia,tja...

 

[Alwin]

How does your corosync.conf look like?

And did you create the cluster with the following?

pvecm create CLUSTERNAME -bindnet0_addr 10.10.10.1 -ring0_addr 10.10.10.1 \
-bindnet1_addr 10.10.20.1 -ring1_addr 10.10.20.1

 

[Thomas Jagoditsch]

How does your corosync.conf look like?

And did you create the cluster with the following?

pvecm create CLUSTERNAME -bindnet0_addr 10.10.10.1 -ring0_addr 10.10.10.1 \
-bindnet1_addr 10.10.20.1 -ring1_addr 10.10.20.1

hi alwin, thx for looking into this.

first i created the cluster without bindnet and ring addresses but before adding the first node i changed /etc/corosync/corosync.conf according to https://pve.proxmox.com/wiki/Cluster_Manager section "RRP On Existing Clusters".
but as said above its a testcluster only. so i recreated/reinstalled all machines on the weekend and created the cluster with the same command that you wrote above.
this time the whole thing worked without any problem.
i must have made some typo i guess.

one thing that confuses me even in the new installation is the existence of 2 corosync.conf files ...

root@test0:~# ls -la /etc/pve/corosync.conf /etc/corosync/corosync.conf
-rw-r--r-- 1 root root     698 Nov 13 10:36 /etc/corosync/corosync.conf
-rw-r----- 1 root www-data 698 Nov 13 10:36 /etc/pve/corosync.conf

... and on my plain new installation they dont seem to be linked.
could you elaborate which one is in use for which purpose ?

tia,tja...

 

[Alwin]

Basically both, the /etc/pve/corosync.conf is copied to /etc/corosync/corosync.conf on change, as on startup the pmxcfs (/etc/pve/) is not available as it needs corosync to be in sync with other members of the cluster. So, if you are doing changes to the /etc/pve/corosync.conf, then they are picked up immediately, whereas the changes in /etc/corosync/corosync.conf are local.

See man pvecm for more information.