In real world, things not always running fine.
For example. When you execute STOP command, sometime the result is not Success, then your next DESTROY command will not success too (because you can not delete a running VM). That is why Proxmox API give us a return value for each command.
Another...
I don't think so. In our network, we use EXISTING Mac from lists rather than generate new one.
We need to use SAME Mac address because we binding it to IP according to DHCP.
You don't have to check each return value, but the status needed to make sure the FINAL result.
For example: you need to check if the VM stopped successfully after STOP command, otherwise CREATE command will be fail (because VM ID already exist).
The GUI should work fine. Maybe you click on wrong UI?
There are 3 Firewall enable location ( +1 on Network dvice ). This option is on VM section, not on Datacenter or Nodes.
According to:
http://pve.proxmox.com/wiki/Multicast_notes#Linux:_Enabling_Multicast_querier_on_bridges
So, multicast will work even Router/Switch don't support, right?
You must enable Firewall on VM 101. Mine like this:
# cat /etc/pve/firewall/1000001.fw
[OPTIONS]
enable: 1
[RULES]
IN DROP -p tcp -dport 80
Also, for Qemu/KVM VM you must enable "Firewall" on Network Device (Click VM 101 -- Hardware - Network Device).
You must understand that Proxmox has 3 levels firewall: Datacenter, Node and VM.
CAUTION!!! In Datacenter -- Firewall: you must set INPUT and OUTPUT Policy = ACCEPT, otherwise you will be blocked yourself!!!
Since you need to Blocking port 80 on VM 101, then you must:
1. Enable Firewall on VM...
I have added:
# echo 1 > /sys/devices/virtual/net/vmbr0/bridge/multicast_querier on both Nodes,
# /etc/init.d/networking restart
But omping still not success for multicast. I'm using Linux Bridge (not OVS Bridge).
root@ns501942:~# omping uds-001 uds-028
uds-028 : waiting for response msg...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.