I have reported this issue by e-mail, but I think it is worth tracking it on the ticket system. I got the 2001:1291:200::/48 prefix allocated to my subnet (R10899), but this is the address range from which tunnel prefixes are being allocated, for the brudi01 PoP. From https://www.sixxs.net/pops/ctbc/ : """ This PoP provides tunnels from the following prefixes: 2001:1291:200::/48 of size /64 Subnets are allocated out of: 2001:1291:200::/40 of size /48 """ From https://www.sixxs.net/home/routeinfo/?10899 : """ Subnet Configuration The configuration for this subnet looks like: IPv6 Them2001:1291:200:f::2/64 (T25828) Prefix2001:1291:200::/48 StateEnabled Created2010-03-01 21:39:27 CET Last Modified2010-03-01 22:07:35 CET """ I can ping addresses inside the subnet prefix, meaning that the prefix is really already in use by the PoP. e.g.: http://www.berkom.blazing.de/tools/ping.cgi?STR=2001:1291:200::1 PING 2001:1291:200::1: 56 data bytes 64 bytes from gw-1.udi-01.br.sixxs.net (2001:1291:200::1): icmp_seq=0. time=263. ms 64 bytes from gw-1.udi-01.br.sixxs.net (2001:1291:200::1): icmp_seq=1. time=264. ms 64 bytes from gw-1.udi-01.br.sixxs.net (2001:1291:200::1): icmp_seq=2. time=267. ms 64 bytes from gw-1.udi-01.br.sixxs.net (2001:1291:200::1): icmp_seq=3. time=262. ms 64 bytes from gw-1.udi-01.br.sixxs.net (2001:1291:200::1): icmp_seq=4. time=261. ms (I don't have any host using addresses from new subnet yet, except for my tunnel address, 2001:1291:200:f::2, that is working as expected).