Right, it's a bug at the Xen layer. I'm also not convinced that "the xen startup scripts are giving us an 'its back' message too early" because it's consistently only the first packet that's a problem, it's not time-based.
Right, it's a bug at the Xen layer. I'm also not convinced that "the xen startup scripts are giving us an 'its back' message too early" because it's consistently only the first packet that's a problem, it's not time-based.