• 5
    • Our product teams collect and evaluate feedback from a number of different sources. To learn more about how we use customer feedback in the planning process, check out our new feature policy.

      Test Crowd for IPv6 compatibility.

            [CWD-837] Officially support IPv6

            Thanks, I'll give it a shot.

            George Lewis added a comment - Thanks, I'll give it a shot.

            Peggy added a comment -

            We've just released Crowd 2.1.0 Beta 4 - which includes support for IPv6

            We would love your feedback on this beta release - but remember, it is a beta, so the usual warnings regarding NOT running it for production systems apply.

            Peggy added a comment - We've just released Crowd 2.1.0 Beta 4 - which includes support for IPv6 We would love your feedback on this beta release - but remember, it is a beta, so the usual warnings regarding NOT running it for production systems apply.

            Support for ipv6 is exceptionally broken. I won't pretend to understand why Crowd has a table with application id, decimal representation of ipv4 address, and binary representation of ipv4 address, but it gets cranky in an ipv6 environment. As you can see from this screen shot, one work around was to jam a row into the db with the ipv6 decimal address and the ipv4 binary address bound to the same interface.

            David Yu's suggestion to add the JVM option (-Djava.net.preferIPv4Stack=true) would only affect traffic outbound (initiated by the jvm), rather than inbound ipv6 requests. That won't resolve this issue.

            We are happy to provide whatever assistance we can when you're ready to address this issue.

            John Eisenschmidt added a comment - Support for ipv6 is exceptionally broken. I won't pretend to understand why Crowd has a table with application id, decimal representation of ipv4 address, and binary representation of ipv4 address, but it gets cranky in an ipv6 environment. As you can see from this screen shot, one work around was to jam a row into the db with the ipv6 decimal address and the ipv4 binary address bound to the same interface. David Yu's suggestion to add the JVM option (-Djava.net.preferIPv4Stack=true) would only affect traffic outbound (initiated by the jvm), rather than inbound ipv6 requests. That won't resolve this issue. We are happy to provide whatever assistance we can when you're ready to address this issue.

            David Yu added a comment -

            I haven't tested this yet but adding this JVM option -Djava.net.preferIPv4Stack=true may help you avoid disabling IPv6.
            http://java.sun.com/j2se/1.4.2/docs/guide/net/properties.html

            David Yu added a comment - I haven't tested this yet but adding this JVM option -Djava.net.preferIPv4Stack=true may help you avoid disabling IPv6. http://java.sun.com/j2se/1.4.2/docs/guide/net/properties.html

            The primary issue with this is host validation of IPv6 addresses. For example, 0:0:0:0:0:0:0:1 and ::1 are equal.

            We'll probably need a new validator that understands how to expand IPv6 addresses so they can be safely compared. We'll also need to understand how IPv6 represents wildcards (ie, its equivalent of CIDR).

            David O'Flynn [Atlassian] added a comment - The primary issue with this is host validation of IPv6 addresses. For example, 0:0:0:0:0:0:0:1 and ::1 are equal. We'll probably need a new validator that understands how to expand IPv6 addresses so they can be safely compared. We'll also need to understand how IPv6 represents wildcards (ie, its equivalent of CIDR).

              pkuo Peggy
              donna@atlassian.com DonnaA
              Votes:
              9 Vote for this issue
              Watchers:
              9 Start watching this issue

                Created:
                Updated:
                Resolved: