I’ve had Vyos virtualized on ESX5.5 for the last two months.
Ever since going production with it, both interfaces eth0 and eth1 loose IP address isolating my network from the internet. Very little errors show in /var/log/messages:
proto: precision =0.108 usec
ntpd exiting on signal 15
zebra[1786] interface eth0 index 2 changed <BROADCAST, MULTICAST>
netplugd[1684] interface eth0 backup up
zenra [1786] interface eth0 index 3 changed <BROADCAST, MULTICAST>
netplugd[1684] interface eth1 backup up
I only recover from this by manual rebooting the VM. Any recommendation or has anyone else run into this issue?
Does the configuration of the router change or do you loose connectivity with the router? What kind of vSwitches do you use? Is it belonging to a VMware cluster? What kind of hardware do you run on? How are your VMware hosts interfaces connected to physical switches etc.
I’m running on VMware 5.5 and do not observe anything like you describe.
The configuration on the router didnt change. Every 3-4 days of the router working fine this problem occurs. I use a standard vswitch.
Running on Dell R720 with Intell I350 Gigabit NICs. I’m using 2 for management VMKernel Ports and the other 6 Virtual Machine Port Group configured with VLANs and connected to HP procurve
All nics are active with LB Route based on the originating virtual port ID.
No error/drops shown on Switch or on show interface eth* physical status.
The VM is setup as a Debian/linux6 (64bit) guest. 2vcpu, 8GB RAM and two Nics.
Which version of Vyos are you running. I’m running 1.1.3
My VMware hosts which are connected to just one logical switch with more than one NIC, I’m using LAG with LACP. For this to work you need to configure the portgroups to load balance based on IP hash. If you are using a HP Trunk (LAG) either with LACP or static you should try to reconfigure your Load Balancing. If you don’t have a LAG you should configure one NIC to be active and on to be standby. You must configure Active/Standby nic on all portgroups.