Need reviewer for 6898234: (dc) Multicast tests fail on OpenSolaris with vboxnet0 adapter
Alan.Bateman at Sun.COM
Mon Nov 9 05:36:21 PST 2009
Christopher Hegarty -Sun Microsystems Ireland wrote:
> Hi Alan,
> I don't have a problem with this change.
> I wonder if some of these failures are dependent on the configuration
> of the guest OSes in VirtualBox? That is, NAT, bridged, Host-only,
> etc... What I'm getting at here is will the IP address always be 0.0.0.0?
It's hard to write good multicast tests that don't require any
configuration. All we can do is filter the "obvious" cases (such as
interface that aren't up or don't support multicast). For these pseudo
interface then it's very possible/likely that they will be plumbed with
IPv4 or IPv6 address, in which case these tests will exercise them. For
default setups where the interface appears to be plumbed with 0.0.0.0
then it's okay/safe to ignore the interface. Any further probing would
require additional support in the API or in the test suite.
More information about the nio-dev