12-15-2013 06:34 AM
I have two Vyatta instances running in VMware. The Vyatta's are configured for a Layer 2 Bridge over OpenVPN. This has been running fine for nearly a year, until a day or so ago.
Site A: has a single subnet, the Vyatta has two Ethernet Interfaces. The Vyatta IS NOT the default gateway on the network. Eth0 faces the inside of the network, Eth1 connects to the ISP. Vtun0 is up and bridged to Eth0.
Site B: has several subnets, no overlaps. There are several Ethernet interfaces connected to various VLANs. Again, no overlaps. The Vyatta is behind a firewall that allows Site A's Vyatta to connect. Eth4 is bridged to Vtun0.
This has been working great for several months. Less than 24 hours ago, my users started calling and reporting that their servers are down, they cannot access the Internet, etc. Running ping -t to the hosts results in 13-15% packet loss.
I run through the switches, the firewall, the ISP router, the virtual networking in VMware. All looks good. I decide I'm going to shutdown all VMs and bring them up a few at a time. As I'm bringin up the VMs everything looks good. When I bring up the Vyatta, everything goes to crap again. I shut it down, everything's good. I disable Eth0, everything's good.