Hello girls and guys!
I recently updatet to Proxmox 3.6-4 and experience a problem that I had believed solved allready in Proxmox 3.2 . I'm dealing with a VZ Container that holds two network interfaces in bridged mode. Each interface refers to a different IP range. E.g. interface veth101.0 is eth0, mac auto, vmbr0 (192.168.20.0) whereas veth101.1 is eth1, mac auto, vmbr1 (192.168.25.0).
The problem occurs when a machine of vmbr0 tries to connect to eth1 or the other way round. The connection becomes instabile. One can test that e.g. with an SSH sesssion connecting to the interface in the other IP-address range. Max. stability lasts around one minute, then the session freezes and must be restarted.
I had a similar problem when using venet adresses in Proxmox 3.2 but changing to bridged network adapters then solved the problem. Now in the new version the problem is reoccuring on the other inteface type. Strangely only with VZs and not on KVMs. So this is what I tried already:
Activating IP-Forwarding:
echo 'net.ipv4.ip_forward=1' >> /etc/sysctl.conf
echo 'net.ipv6.conf.all.proxy_ndp=1' >> /etc/sysctl.conf
Changed following entry in /etc/vz/vz.conf
# Controls which interfaces to send ARP requests and modify APR tables on.
NEIGHBOUR_DEVS=all
Source for this idea under http://wiki.euserv.de/index.php/Manual_Einrichtung_Proxmox/en#Creating_an_OpenVZ_container
Passage:
Configuration OpenVZ routed
Hmm, formating this text semms to exceed my skills. Sorry for that. Anyway, I even restarted the server and could unfortunatly not observe a change in this behaviour. SSH sessions still last only about a minute from one network segment to the other.
In case you ran into similar trouble your help is greatly appreceated. It would also help me a lot to know which services I can restart while trying to solve this issue instead of restarting the whole server.
Thank you very much and kind regards, MisterIX.
I recently updatet to Proxmox 3.6-4 and experience a problem that I had believed solved allready in Proxmox 3.2 . I'm dealing with a VZ Container that holds two network interfaces in bridged mode. Each interface refers to a different IP range. E.g. interface veth101.0 is eth0, mac auto, vmbr0 (192.168.20.0) whereas veth101.1 is eth1, mac auto, vmbr1 (192.168.25.0).
The problem occurs when a machine of vmbr0 tries to connect to eth1 or the other way round. The connection becomes instabile. One can test that e.g. with an SSH sesssion connecting to the interface in the other IP-address range. Max. stability lasts around one minute, then the session freezes and must be restarted.
I had a similar problem when using venet adresses in Proxmox 3.2 but changing to bridged network adapters then solved the problem. Now in the new version the problem is reoccuring on the other inteface type. Strangely only with VZs and not on KVMs. So this is what I tried already:
Activating IP-Forwarding:
echo 'net.ipv4.ip_forward=1' >> /etc/sysctl.conf
echo 'net.ipv6.conf.all.proxy_ndp=1' >> /etc/sysctl.conf
Changed following entry in /etc/vz/vz.conf
# Controls which interfaces to send ARP requests and modify APR tables on.
NEIGHBOUR_DEVS=all
Source for this idea under http://wiki.euserv.de/index.php/Manual_Einrichtung_Proxmox/en#Creating_an_OpenVZ_container
Passage:
Configuration OpenVZ routed
Hmm, formating this text semms to exceed my skills. Sorry for that. Anyway, I even restarted the server and could unfortunatly not observe a change in this behaviour. SSH sessions still last only about a minute from one network segment to the other.
In case you ran into similar trouble your help is greatly appreceated. It would also help me a lot to know which services I can restart while trying to solve this issue instead of restarting the whole server.
Thank you very much and kind regards, MisterIX.
Last edited: