Sure - there is still 16.3.x which seems to be overall, quite solid and I think will be around for some time yet. But going forward the feature has been removed.
However the real takeaway from this change is that Cisco no longer see the Converged Access model as we know and as they have been marketing it, as the way forward for everyone.
I tried CA in a lab/nonprod environment and it was a disaster in terms of a brownfield deployment because I didn't have CA switches to connect to every single AP all at the same time. It might be OK for a greenfield deployment but the transition from existing to new CA is a nightmare all for relatively little gain. I had to keep my existing WLC anyway so I saved nothing, gained nothing, but added complexit, wasted time and it cost license fees. FlexConnect works well for most people and CA just seemed and felt like a bad hack - although the concept sounded good. Additionally CA code seemed to lag way behind that of the AireOS code in terms of features, AP support and the CA code seemed to be constantly in a state of catch-up. I had to wait for switch firmware updates to pick up wireless features and new AP support, for example. This meant the switching team were indirectly dictating what and when I could deploy a new generation AP.
In addition in a corporate environment as an engineer I then had a double blow of determining if an IOS-XE upgrade on my switch was also going to fix switch bugs but break my wireless environment. This made switch firmware upgrades very high risk and the process for rolling back was messy because I would have had to roll back both. (Yes there was talk about the capability of splitting out packages and being able to upgrade independently, but there hasn't been a single case where Cisco actually released separate packages to do this - so in my mind it was only ever marketing fluff).
In my view Cisco seem to have realised now that CA was a mistake and it won't take off. Perhaps if they take a step back they may realise why and what the shortcomings were with it in terms of real world use cases.
What may work well is if they can integrate and decouple the wireless code from AireOS into the Cat3k, but as a VM. Then we would get the best of both worlds and they would be onto what I think would be a winning idea.
Apparently there is something in the works going forward with regards to CA. I am not privy to what it is, but whatever it is I hope they get it right this time. The cat3k is a great platform with lots of capabilities but wireless to date has not been one of the more useful of them.
... View more
.jpg "VIP Engager"){kind=icon}
.jpg "Hall of Fame Guru"){kind=icon}
