VM isolated

[igurvich]

I've installed a Ubuntu 16.04 server vm on a node in my cluster but for some reason it's isolated from everyone and I can't ping it even from the node it's located on. when I run tcpdump it shows that the ARP request for the host is looping and can't find it.
Node IP: 172.16.17.120
VM IP 172.16.17.148

ip addr


/etc/network/interfaces


VM network interface config


Also when I start running the vm I receive an error message saying "failed to run vncproxy" , but the machine starts fine. Is there a connection?

EDIT: I'm still trying to troubleshoot and created another VM inside the same node and these two can ping each other. Trying to tcpdump on the interface on the node where the vm is located i see no traffic from these nodes despite them being bridged to this interface.

 

[dcsapak]

how does your host network config look like ?

 

[igurvich]

how does your host network config look like ?

I've made a few changes but here is the topology of the network I've built:


the 192.168.17.0/24 is a host only network and the 172.16.17.0/24 is NAT.


The VMware machine config of the host


ip link output


/etc/network/interfaces


from the web portal


ip route output

 

[dcsapak]

but you gave the vm an ip in the 172.x.x.x range not 192.x.x.x (so not like in your picture)?
also why the pointtopoint entry in the vm network config ?

 

[igurvich]

but you gave the vm an ip in the 172.x.x.x range not 192.x.x.x (so not like in your picture)?
also why the pointtopoint entry in the vm network config ?

Whoops forgot to upload the updated info for the VM:

 

[dcsapak]

Whoops forgot to upload the updated info for the VM:

ok so what does now work and what does not?

 

[igurvich]

So after rebuilding my entire network and playing around with options I've changed the interface model from virtIO to realtech. Seems to have fixed the issue.