Question

Expand Cluster - IPv6 needed in CE allways?


Userlevel 2
Badge +12
Tried to expand my Cluster from 3 to 4 nodes. Get the following error:



Next try was made like one knwon blogger explains with ipv6 pinging etc.
Same result.

Found and chinese/japanise articel with working of redirecting the local proxy for discovery.

Any Ideas how to expand a 3 to 4 node cluster with a windows or mac management client?

Thanks

This topic has been closed for comments

16 replies

Userlevel 2
Badge +12
These Workaround with ncli doesnt work too:
http://virtual-hike.com/add-new-node-nutanix-cluster-using-cli/
Userlevel 2
Badge +12
These Articel describes the same problem and a solution but i could not translate the content so far...
https://qiita.com/albatross/items/e289adabdb49cfd7f4a5
Userlevel 7
Badge +34
Hi @frsbeckum

This is for CE correct? I'm moving to the CE forum
Userlevel 2
Badge +12
yes. CE in Homelab. In Production i use more then 4 nodes....
Userlevel 5
Badge +9
Hi,
Do you get the same error in ncli as in the UI?
I've looked around the forum topics and found a few similar issues like these:
https://next.nutanix.com/discussion-forum-14/expand-cluster-from-3-to-4-nodes-ce5-0-18282
https://next.nutanix.com/discussion-forum-14/expand-a-ce-cluster-problem-19868
Although there seem to be subtle differences between them so I'm not sure if this helps.
Userlevel 2
Badge +12
Hi,
shawn touch iso hint seems to be good. But i see one problem on my new node:


How could i set the hypervisor ip (192.168.10.71) on these new node?
Userlevel 5
Badge +9
There are some procedures in the Advanced Setup guide, but I did not check if this really fits your case:
https://portal.nutanix.com/#/page/docs/details?targetId=Advanced-Setup-Guide-AOS-v55:Manual-IP-Address-Configuration-Methods
Userlevel 2
Badge +12
no. sorry
Userlevel 7
Badge +34
You still having issues @frsbeckum ?
Userlevel 2
Badge +12
a lot of issues...no ip of hypervisor no expandig. see screenshot
Userlevel 7
Badge +34
Do you have anything to add @andrew_ct @Primzy ?
Userlevel 5
Badge +9
Hi,
if I correctly understand the issue ... adding the node with
ncli cluster configure-node node-uuid="node_uuid" [hypervisor-ip="hypervisor_ip" ] ...
fails due to missing IPs? What does the command report?
As for changing the host address, I would normally try by updating /etc/sysconfig/network-scripts/ifcfg-br0 (as stated in https://portal.nutanix.com/#/page/docs/details?targetId=AHV-Admin-Guide-v51:ahv-acr-host-ip-address-change-t.html), but if I correctly understand your answer, this does not work for you?
Userlevel 2
Badge +12




Host Network seems ok on KVM Host:



Result after Genesis Restart in CVM and Discover on Cluster Node:

Badge +6
I ran into a similar issue at one point, as i remember it, it was related to missing VLAN settings on the switch (cluster had many VLANS configured, but the new port on switch was not configured for anything but default VLAN)
and i setup IPMI to a 192.168.0.5 address which made not real sense, but it allowed me to add the new node.
Userlevel 7
Badge +34
Hi @frsbeckum did you see the reply from @Rasmus - might be helpful.
Userlevel 2
Badge +12
Intel Skull Canyons have no IPMI Port!
Hypervisor and CMVs are in the same Network. No VLAN Tagging is used!