We were pretty excited here when version 8.2 of the ASA OS was released to the public a few weeks ago. Not only was IPv6 failover to be supported in the release (per Cisco TAC — see previous entry), but as I perused through the release notes I saw several other important IPv6 enhancements: IPv6 support in ASDM version 6.2, IPv6 support in transparent mode and IPv6 support for IPS. Interestingly, the release notes did not mention something as important for enterprise IPv6 adoption as IPv6 failover support, so I decided to dig a bit deeper before diving into an upgrade. Sure enough, in the “Failover System Requirements” section of the 8.2 CLI Configuration Guide: IPv6 failover is not supported in Release 8.2(1). This was a disappointing find, but I decided to remain optimistic and maintain the possibility that maybe I’d just run across a documentation error. Going back to the source, I opened another TAC case (SR 611470841). The tech was extremely helpful, informing me that while IPv6 failover support was on Cisco’s roadmap, there was no specific release targeted for inclusion of this "feature".
Basically, since we are running a failover pair in our datacenter, IPv6 is still not an option for us on the ASA. I find it strange that Cisco would devote development time and resources to the IPv6 enhancements listed in the release notes while neglecting critical functionality like IPv6 failover support, the absence of which precludes the possibility of ANY ASA IPv6 deployment in a failover environment. Even if I deem IPv6 as non-critical traffic (at this point) and do not require IPv6 failover capabilities, the lack of support for IPv6 in the failover configuration (or at least the ability to ignore IPv6 commands in the config that is synced to the standby unit), ensures that configuration of IPv6 on my failover pair will result in unpredictable behavior from the devices on my network.
As the IPv4 deadline draws near, enterprises interested in testing and deploying IPv6 services may begin to look to other vendors for the functionality they require. In our case, a demo Juniper SSG (that we’d had no real intention of deploying) is now running parallel to our ASA failover stack, and has been running flawlessly since we deployed it.