@techreport{ybai-ipv4v6-transition-practice-in-openstack-00, number = {draft-ybai-ipv4v6-transition-practice-in-openstack-00}, type = {Internet-Draft}, institution = {Internet Engineering Task Force}, publisher = {Internet Engineering Task Force}, note = {Work in Progress}, url = {https://datatracker.ietf.org/doc/draft-ybai-ipv4v6-transition-practice-in-openstack/00/}, author = {Yi Bai and Congxiao Bao and Kevin Yin and Xing Li}, title = {{IPv4/IPv6 Transition Practice in OpenStack}}, pagetotal = 11, year = 2014, month = sep, day = 21, abstract = {OpenStack is a free and open-source software cloud computing platform. It is primarily deployed as an infrastructure as a service (IaaS) solution. However, OpenStack is designed mainly for IPv4, it internally uses {[}RFC1918{]} addresses and heavily relies on NAT to map RFC1918 addresses to public IPv4 addresses known as floating IP addresses for the external access. Due to the different nature of IPv6 and IPv4, the IPv6 support for the OpenStack is still in the early stage. In this document, two mechanisms are presented to provide IPv4/IPv6 dual stack external access for the OpenStack, one scenario is internal IPv4 and uses stateful IPv4/IPv6 translator for the external IPv6 access, and another scenario is internal IPv6 and uses stateless IPv4/IPv6 translation for the external IPv4 access. Both mechanisms have been deployed in CERNET and providing services to the global IPv4/IPv6 Internet.}, }