Containers losing IP Adress

[JannikJung0]

Hey all,
For some time now I have a weired problem with my proxmox containers.
I am running about 30 Fedora27 containers across multiple servers, and some of them simply lose their IP Adress and therefore breaking networking on them.

The time when the containers lose their connection varies and this problem is not bound to any single server.

Did someone experience something similiar or how could I go forward debugging this problem?

Any help is appreciated!

Thanks!

 

[Stoiko Ivanov]

Any relevant information in the journal (of the host and of the containers) at the point where they lose the ip address?
Take a look of the output of ip link , ip addr, ip route in the container

 

[JannikJung0]

My problem is, that I do not exactly know when they lose the ip. Furthermore, when that happens, the log gets spammed with telegraf messages, not being able to send to the influx server, making it impossible to see any other problems.

Nonetheless, here is the output of ip link:

[root@salt ~]# ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
143: eth0@if144: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
    link/ether be:32:7a:d2:9b:1b brd ff:ff:ff:ff:ff:ff link-netnsid 0
[root@salt ~]#

IP Route doesn not output anything at all.

I also checked the log on the host, and this message keeps showing up.
I have a LACP Bond setup, but I feel like the switch is acting kind of weird, so I will replace it this afternoon and check whether it dissappears.

vmbr0: received packet on bond0 with own address as source address