The scope of this solution is to help customers understand what causes the detection of duplicate Address when running IPv6 tests and how to change the configuration to bypass this issue.
When running an IPv6 test, a Duplication Address Detection is performed from the hosts at the beginning of the test, to see if their IPv6 address is used from another host. The hosts will broadcast their Link local Address, which is composed by the prefix "fe80:0000:0000:0000" (64 first bits) and the latter 64 bits are the ones of its static global address, as configured in the Subnets.
If when creating our different Subnets, we use different 64-bits prefix but the same 64-bits suffix, we will have a conflict of Link Local addresses resulting to an error message "Duplicate address detected: fe80::xxxx. Host disabled". If in our topology, 4 client ports are connected through a switch, DAD messages sent from port 1 will be received by the other 3 ports which consider the addresses in DAD messages duplicate. As a result the hosts in 3 ports won't be started and only one Subnet will come up. 2001:DB8:1:AA::10 - 2001:DB8:1:AA::FF 2001:DB8:1:AB::10 - 2001:DB8:1:AB::FF 2001:DB8:1:AC::10 - 2001:DB8:1:AC::FF 2001:DB8:1:AD::10 - 2001:DB8:1:AD::FF
To resolve this issue, you should differentiate the last 64 bits in the IPv6 Subnets tab, which will result to the creation of different Link Local Addresses, the DAD procedure will find no Duplicate Addresses and all the subnets will come up. 2001:DB8:1:AA::A:10 - 2001:DB8:1:AA::A:FF 2001:DB8:1:AB::B:10 - 2001:DB8:1:AB::B:FF 2001:DB8:1:AC::C:10 - 2001:DB8:1:AC::C:FF 2001:DB8:1:AD::D:10 - 2001:DB8:1:AD::D:FF