0

I have a two node ( VMs ) setup running keepaliveD. The problem is that due to only ability of managing the VMs from command line, I can't demonstrate how the network break between both node-A and node-B will behave ( if I would have a control, I could have simulated it by disabling the network on any node though, if there any other procedure, please suggest me )

Idea is to use a 2 node cluster with keepaliveD only. what happens when the network goes down is still to be discovered

Any recommendations?

Nasir Mahmood
  • 21
  • 2
  • First, when operating a VM, you should always have some backup path to reach your machine. In most cases this means having a virtual KVM or emergency console. If this is not given, you could try to set this up using 2 local VMs in a VirtualBox or any other virtualization of your choice, then disconnecting the interface(s). If this is not possible either, you may play with a firewall restricting all (except SSH) or only VRRP traffic. This would eventually lead to keepalived on both sides enabling the IP, causing a flipflop, which should not matter as the firewalled machine will be encapsulated. – M. Schmidt May 10 '20 at 10:27
  • Schedule a reboot of node-A (shutdown -r). Shut down its network temporarily (ifdown). Keepalived should activate node-B immediately. Check if everything is OK and wait for node-A to come up again. – Gerard H. Pille May 10 '20 at 15:19

0 Answers0