Thread Info
  • State Suggested Answer
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 4 subscribers
  • Views 8793 views
  • Users 0 members are here
Options
Suggested
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

HA cluster not working

Anth

I'm currently attempting to cluster two virtual UTM boxes.

utm1 - 192.168.106.88

utm2 - 192.168.106.178

 

Setting up utm1 as node 1 I enter the shared secret on both boxes and set them both to use eth2. They both get 192.168.106.88 which is what happens when I have set up other clusters. But I can't access the the web interface via https://192.168.106.88:4444

 

Does this sound like an arp issue, as if both nodes 'think' they are active? running ha_utils shows one as the master and one as the slave...

 

I killed HA and only set it up on utm1 only. Usually I would see H/A running with 1 node in the web interface, and ha_utils shows that there is one node, but again I lose the ability to ping 192.168.106.88 or access the web interface. In the past this has usually worked O.

Anyone seen this sort of behaviour before?

Cheers

Anthony  



This thread was automatically locked due to age.
  • 0

    Hi Anthony,

    Eth3 is the default HA sync interface for the UTM, if you are trying to sync from eth2 then make sure to change the sync interface to eth2 on both the appliances. Alongside, both the UTM shouldn't get a similar IP address. Please show us a picture of the output for command "ha_utils" in the shell. Also, show us the high-availability.log if you are able to SSH to the Master UTM.

    Thanks

  • 0

    Anthony, properly configured, the Master is the only VM that should be configured.  When you activate the second VM, it should see the Master and the Master should configure it.  I would not have expected to see an addressing scheme of 192.168.106.88/178 on the sync interfaces - are those IPs that you put on Internal interfaces?  The easiest would be to create new VMs with at least four NICs (VMXNET3 preferred with ESXi) so that you can let eth3 be the sync NIC.

    Cheers - Bob