On 06/21/2020 02:17 PM, Oliver Dzombic wrote:
Hi,
we are currently evaluating openvz 7 and run into the following problem:
Normally by executing:
prlctl start $UUID
the system will create two arp entries:
? ($IP) at <from_interface> PERM PUB on virbr0
? ($IP) at <from_interface> PERM PUB on br0
( we are using host routed network )
But here it populates:
? ($IP) at <from_interface> PERM PUB on virbr0
only.
So we can ping the hostmachine, as well as the hostmachine can ping the
container IP.
But the container can not reach the outside, because the outside has no
arp entry to return the requests.
How to resolve this ?
Hi Oliver,
it works for me, so please file a bug at bugs.openvz.org and post there full
commands with outputs for reproduction.
On my node:
[root@sbf7 ~]# vzctl set 100 --ipadd 10.10.0.34/24
Adding ip address(es): 10.10.0.34
vz-net_add WARNING: ! /sbin/arping -f -c 1 -q -w 1 -I br0 10.10.0.34 FAILED
arping: Device venet0 not available.
arping: Device venet0 not available.
WARNING: Settings were not saved. On next start the original values will be
applied. Use --save to save the settings in the configuration file.
[root@sbf7 ~]# arp -n | grep 34
10.10.0.34 * <from_interface> MP
docker0
10.10.0.34 * <from_interface> MP
virbr0
10.10.0.34 * <from_interface> MP br0
Hope that helps.
--
Konstantin
_______________________________________________
Users mailing list
Users@openvz.org
https://lists.openvz.org/mailman/listinfo/users
