We have two nodes with proxmox ve 3.4. There are two subnetworks on claster: netA - 192.168.x.x/24 for CT's that no need real ip-address ; netB - for CTs that need real ip. Nodes has ip-addresses from netA by them selfs. Also nodes have proxy_arp enabled like this:
All VMs and nodes are in the same vlan. So cisco router has one address from netA to route netA and has spcial route to netB directly to this vlan to route netB.
All works fine. I create CT with address from netB and it works. But when I do live migration CT with netB address CT stops respond until 'clear ip arp <CT ip>' on cisco router. There is no such problem when I do live migration CTs from netA.
I think node should send some kind of arp of CT after migration and router would change its arp entry for migrated ip-address. I have monitored arp traffic by tcpdump on target node -- there was no any traffic for/from migrated ip.
How can I solve this problem for netB ?
Code:
# cat /etc/sysctl.d/vzctl.conf | grep -v "#"
net.ipv4.ip_forward = 1
net.ipv4.conf.default.forwarding=1
net.ipv4.conf.all.proxy_arp = 1
net.ipv4.conf.default.proxy_arp = 1
net.ipv4.conf.all.rp_filter = 2
kernel.sysrq = 1
net.ipv4.conf.default.send_redirects = 1
net.ipv4.conf.all.send_redirects = 0All VMs and nodes are in the same vlan. So cisco router has one address from netA to route netA and has spcial route to netB directly to this vlan to route netB.
All works fine. I create CT with address from netB and it works. But when I do live migration CT with netB address CT stops respond until 'clear ip arp <CT ip>' on cisco router. There is no such problem when I do live migration CTs from netA.
I think node should send some kind of arp of CT after migration and router would change its arp entry for migrated ip-address. I have monitored arp traffic by tcpdump on target node -- there was no any traffic for/from migrated ip.
How can I solve this problem for netB ?
Last edited:
