Alcatel-Lucent (previous CloudAve coverage), a leading vendor in mobile and networking space, last week announced Alcatel-Lucent Cloudband, set of network services offered through the cloud. Networking is a significant part of cloud infrastructure but it has attracted much less media attention than its counterparts, compute and storage. Part of the reason is somewhat of a muted interest from the traditional networking vendors towards the cloud paradigm initially. Off late, this trend has been changing and we are seeing vendors showing interests even on open source cloud platforms like OpenStack. With this move, Alcatel is trying to jump past other vendors in the space and they are trying to target the telecom service providers. It is an interesting strategy and I am particularly interested because of my interest in the idea of federated clouds.
One of the biggest concerns quoted by enterprises about public cloud services is the availability and reliability. They are worried whether their critical applications will be available whenever they need it and whether it will offer the same reliable performance as their internal data centers. Another big concern is about security. In the first iteration of cloud services (I use this terminology to differentiate cloud services before the onset of federated cloud ecosystems and I am least interested in waging a social media war on the terminology. Feel free to dismiss the term after you read it here. It is just for my convenience), networking didn’t get the same lever importance/interest like compute or storage. The dominant public cloud providers at that time were least interested in getting tech media excited about the networking part. This resulted in enterprises showing a meh interest on cloud services. However, it is changing now and we are seeing a new breed of cloud service providers, some from the telecom sector, who emphasize on network performance as a differentiating factor. More and more service providers are realizing that they need to focus more on the networking part if they want to get into the enterprise market. This realization is opening up a huge market for the networking vendors.
Alcatel-Lucent is trying to attack this pain point faced by service providers in the federated cloud era. They call Cloudband to be the foundation for a new class of ‘carrier cloud’ services. They position carrier cloud services as something that will enable communications service providers to bring the benefits of the cloud to their own networks and business operations, and put them in an ideal position to offer a new range of high-performance cloud services to enterprises and consumers. Sure there is definitely a bit of marketing pluff here but the idea is to move the idea of networking from the physical switches and networks to services that can be tapped from the cloud. With this move Alcatel-Lucent is trying to jump ahead of the traditional networking vendors to get the mindshare of service providers. Whether it will work or not is not something I can predict now but it is an interesting offering.
I call it interesting because with services like these, the role of CDN may get diminished. For example, let us imagine Netflix using a federated ecosystem of cloud providers instead of Amazon EC2. Netflix can then take the processing of movies close to users by tapping a provider who is physically close to the user than processing it on Amazon data centers and then tapping the CDNs for delivery. I may not be 100% technically sound with this example and I might miss out some technical nuances here but I think Netflix may be able to offer better performance at a lower cost if they take the processing of movies closer to their users than doing it at a centralized location. Don’t pick hole on my Netflix example but I am trying to highlight a potential use case here. Recently, Datasift talked to Paul Miller of Gigaom Pro (Subscription Required) about how they cannot trust public cloud providers because of the network performance issues they face in delivering real time monitoring and updates of social media information. I am now aware of all their technical needs but I think Alcatel-Lucent Cloudband coupled with cloud providers in the federated ecosystem can offer a solution that can meet their needs by taking the processing close to their users. These are some of the use cases I could think of while learning about the new Cloudband offering. I get a feeling that it will come handy for the service providers in a federated cloud ecosystem where performance and reliability is also a possibility with other cloud features. I would like to hear the thoughts of practitioners about this offering. Feel free to jump in.
Related articles
- Alcatel Promises Better Clouds for Carriers (pcworld.com)
- A ‘Carrier Cloud:’ Alcatel-Lucent’s Bid to Compete with Amazon (readwriteweb.com)
- Roundup: VCE, SAP, Cisco, Juniper, Alcatel-Lucent (datacenterknowledge.com)
- Getting China Hardwired (dailyfinance.com)
The problem with CloudBand is that I have yet to see any detail on exactly what it *does* (or will do, when they release it), so this makes the use cases and value proposition somewhat hard to judge. Fortunately I have a meeting scheduled with Alcatel on this topic, so hopefully all will be revealed soon.
I agree. I am speculating at this point based on my discussion with them. I can only say for sure when they actually release the service.
thanks for discussing what we’re doing with CloudBand. what we’re doing is managing the network (wide area) together with distributed cloud nodes (built inside the network) in one platform. this then enables us to optimize both networking resources and to choose the optimal cloud placement too meet the apps SLAs (such as latency, bandwidth, gitter, HA, etc..).
Additionally, as Krish wrote, lower TCO for chatty applications when they are placed at the edge, a la CDN.
Dan – happy to dive deeper with you.
Dor