OpenStack is MANO? No…No, It Is Not
There was an article on Tech Target about a year ago with the question, “Are OpenStack and NFV MANO interchangeable?” A colleague stumbled on this recently and asked my opinion on the topic, how true it may have been then and how true it may be now.
I can see where people that don’t get what MANO is could ask this question. The short answer is that OpenStack is a part of MANO, specifically the VIM. Of course, as this is a blog post, I am going to give a little longer answer than that.
The ETSI MANO definition itself is not completely sufficient to really handle what is needed in the real world, and it covers a lot more than OpenStack alone does. ECOMP is much closer to that, but it is really just an extension of ETSI MANO concepts, so that is not necessarily a blow to the ETSI standard.
Can OpenStack be extended to be MANO? Of course it can. There are projects like Tacker and others that are trying to go in that direction but they are miles away from being anywhere near the commercial options from NetCracker, Cisco and even Juniper. These projects are outside of what people consider to be the core OpenStack application. There are ~50-ish projects out there in the OpenStack ecosystem that cover so many areas, so there are tons of things that you wouldn’t normally consider if you mean “cloud infrastructure management” when you say OpenStack. We at Itential could even extend Pronghorn to replace your entire OSS/BSS stack, including Billing…but that wouldn’t be very smart and isn’t something that should be entertained.
With regard to the OPNFV efforts not focusing on the MANO stack higher than the VIM, that is not entirely true a year later either. It was for some time, but in recent months we have seen the Open Source MANO (OSM) project and the Open-O project kick off to provide OPNFV components to fill in those gaps. OpenStack will be a key piece of these efforts because it is the VIM component in MANO for OPNFV.
Many of the above items were also true, or in some cases even MORE so a year ago. This was, and is for those that may still believe OpenStack is the end-all/be-all for MANO, an over selling of OpenStack. It is a fantastic tool for what it does. For smaller implementations, maybe on an Enterprise versus Service Provider scale, the projects like Tacker will become suitable tools for other layers of MANO. However, for large implementations, especially Service Provider, separate MANO components are the way to go and probably will be for the foreseeable future.
If anybody has any interest in discussing further, disagrees and would like to discuss that further, feel free to Contact Us…