cloudstack-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris <>
Subject XenServer 6.2, mgmt nic bonding, bridge mode and cloudstack 4.4.2 error
Date Sat, 21 Feb 2015 10:57:15 GMT
G'day all,

Quick run down of the configuration and architecture I've got;

1 x management server (mgmt) running Centos 6.6 x64, cloudstack management
installed, one nic
2 x xenserver 6.2 (xen1 & xen2) no patches installed, 4 nics
1 x nas (nas) sharing primary and secondary nfs, 2 nics

On the xenservers, I've created a pool with xen1 being the master, then
I've created two network bonds - bond0 for eth0 & 1, bond1 for eth2 & 3 .
Management interface has been assigned to bond0. labels are cloud-private
bond0 , cloud-public bond1 .

At this point, i made no changes to the xe-switch-network-backend, leaving
it on vswitch.

When adding pool (using xen1 ip address) to cloudstack using the basic
setup wizzard, the below error occurs when attempting to add host:

Cannot transit agent status with event AgentDisconnected for host 1,
mangement server id is 94980626781222,Unable to transition to a new state
from Creating via AgentDisconnected

After some research, I saw notes on the need to setup bridge mode by using
the command xe-switch-network-backend bridge . Also to edit the bridge
settings in sysctl.conf by using the following command;
sed -i -e '/net.bridge.bridge-nf-call-iptables/
c\net.bridge.bridge-nf-call-iptables = 1' -e
c\net.bridge.bridge-nf-call-arptables = 1' /etc/sysctl.conf

After performing this on both my xenserver nodes and rebooting - they no
longer have any network connectivity, they can't ping each other nor
anything else on the network. It's not until I switch it back to vswitch
and reboot that they resume on the network. However this configuration
isn't supported by CloudStack.

Any thoughts on how I can fix this?


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message