----- Original Message ----- Hi Heiko, I think "site local address" is used mainly in IPv6 test plan cases of EAP. But NOTE: "site local addresses" concept is obsoleted in RFC too (and RFC 3879 refuses them), also I got reply from Mike (mharvey) last week with IPv6 review feedback from NTT where this is also mentioned. Further info is in my comment sixth in [1] article. [1] https://docspace.corp.redhat.com/docs/DOC-71739 Cheers, Pavel

The site-local-address attribute in the management API's interface resources is used to select an interface based on whether or not java.net.InetAddress.isSiteLocalAddress() returns true. For an IPv4 address that method returns true if it's in one of the IPv4 private address spaces. For Inet6Address it's public boolean isSiteLocalAddress() { return ((ipaddress[0] & 0xff) == 0xfe && (ipaddress[1] & 0xc0) == 0xc0); } (At least it is in my current JDK 6 install.) I think it's an important criteria to have. The name could perhaps be better if the whole notion of a site local IPv6 address is under debate in RFC land. However, if this topic is still being hashed out in the specs, IMO we're better off tying ourselves to the JDK and making this something where we simply expose what the JDK provides. We can call it something else, but it should be documented as simply checking whether InetAddress.isSiteLocalAddress() returns true. On 11/15/11 4:35 AM, Heiko Braun wrote: -- Brian Stansberry Principal Software Engineer JBoss by Red Hat