Closed RFRIEDM-Trimble closed 2 years ago
Hi @RFRIEDM-Trimble ,
Thanks for your detailed report! We have take a look at it and everything seems alright with it. Just a couple of things:
FF1E::FFFF:239.255.0.1
is not a format we currently support for constructing or parsing. I'd suggest using something like the Locator constructor, and then set the address with one of the IPLocator::setIPv6
functions. Mind that the 239.255.0.1
part would need to be translated into hex double wordsA PR on your part would indeed be most welcome!
Again, thanks for the interest and the thorough report.
No problem.
Yep; I'll adjust the address format to be compatible the API's and submit a PR soon.
I created it with FF1E::FFFF:239.255.0.1
which expands to ff1e:0000:0000:0000:0000:ffff:efff:0001
and compresses back down to FF1E::FFFF:EFFF:1
I will proceed and close this issue as the related PR #1959 was merged several months ago (May 2021) and it was also backported to previous branches (from 2.0.x
onward)
The DDS 2.3 standard only dictates what the IPv4 UDP default address is in section 9.6.1.4. FastDDS does not currently implement a standard IPv6 address.
Thus, I don't think there is a right way to initialize a default IPv6 address multicast locators, but there should be two requirements.
Currently, one address is used here and another here, the latter not being a valid IPv6 address. This ticket is to address that problem.
For IPv4, the standard is correctly followed in the code here
A proposal would be this.
FF1E::FFFF:239.255.0.1
to follow RFC4291 for the default IPv6 address for Fast-DDSflgs
T is set to 1 totransient multicast address
flgs
P
is also set to 1 to indicate multicast address assigned based on the network prefix following RFC3306scop
is set to Global ScopeE
flgs
R
is set to 0While you're at it, might as well put that address in a header file so other applications can use it.
I'm happy to do the PR; just want feedback on the above.
_Originally posted by @RFRIEDMTrimble in https://github.com/eProsima/Fast-DDS/issues/1640#issuecomment-776184596