A survey that Light Reading done among the carriers, clearly showed that there has been a major increase in strategic activity related to cloud services over the past year. More than 85% of telecom operators responded to this survey, said that their companies are either offering cloud services or are taking steps to offer them. Moreover a solid majority believes that telecom service providers have a competitive advantage in providing cloud services, because they control the network and can provide end-to-end SLAs.
Image may be NSFW.
Clik here to view.
As enterprises and consumers access these new cloud based services, connectivity by end users or business locations will mandate aggregated WAN connections into the private or public based clouds. Recent advances in wide area networking has brought Carrier Ethernet to the forefront as the transport mechanism of choice to access both private and public cloud services as well as cloud-to-cloud connectivity.
Carrier Ethernet connections will have to support new capabilities to address the unique needs of connections to cloud based services. Several areas of focus and use cases have been identified by the MEF: See new white paper on Carrier Ethernet for Delivery of Private Cloud Services
- Elastic Ethernet Service attributes
- On-demand bandwidth approaches and
- Network/Cloud/Subscriber Management Orchestration
- It connects the core network with the access and edge network
- It connects the cloud traffic to the end-users
- It also connects between different clouds/datacenters (different locations and different providers)
But there is much more than just connectivity. To ensure end-to-end SLA, various elements need to be controlled and managed in order to provide successful cloud based service:
- Scalable density: Carrier Ethernet connections have evolved from mid-band Ethernet connections (sub 100Mbps) to Gigabit and now 10GigE connections for wholesale, business end users, and for mobile backhaul connections. New Carrier Cloud gateway platforms will enable scale via modular chassis based designs that can scale to support high numbers of both 1G and 10G interfaces which will drive multiple 40GigE connections and 100GigE (future) for the uplinks with a minimum of at least a 1T of total capacity.
- Open Architecture: To ensure rapid deployment and adoption of these new carrier cloud gateways, the equipment architecture should ideally be based on industry standard such as AdvancedTCA chassis. These standard chassis can scale from 1RU up to 16RU slot capacity, allowing operators to rapidly match density needs while adding new applications capabilities via standard AdvancedTCA ecosystem blades and line interface cards. An ecosystem of AdvancedTCA vendors already support best of breed applications such as firewall, intrusion detection, video optimization, DPI, caching, load balancing, SAN, virtualization, policy management and new SDN based Openflow protocol.
- Efficiency: Carrier Cloud gateways will need to support best in class densities with space, power savings, and extensive Ethernet wire speed OAM to ensure service providers can easily manage and monitor SLA requirements while minimizing CAPEX and maximizing OPEX savings for low overall TCO.
- Flexible: Several protocol alternatives exist to support end to end carrier Ethernet connectivity. Ideally the carrier cloud gateway will have support for both L2 Ethernet, MPLS, and new SDN supported Openflow protocols to ensure carriers can support interconnection to any customer, carrier, operator, or wholesale broker that may require access to the carrier cloud gateway.
- End-to-end service management and monitoring: MEF identified that there must be smooth management orchestration between carrier network, cloud provider, and subscriber management systems. Additionally operators will likely require multivendor support, as different existing vendors equipment will continue to exist in the operator networks as migration to newer cloud architectures are deployed. As part of the services management, more granular monitoring of the actual end-user flows will increasingly be needed to provide real time information for security and virtualization techniques which will be used to provide secure bandwidth-on-demand and elastic Ethernet capabilities to “tune up or down” service levels on an as-needed-basis.
- Flow based management – the quality of different flows and different services should managed separately to ensure different characteristics per type of service per customer some mechanisms like HQoS exist already in the market today allowing that
- Resiliency - as the network getting more and more complex resiliency is not only about providing alternative to failing network device but alternative to depredated service and resiliency across different domains
- Low Latency – the level of experience user expect should be as high if not higher as the service the users are getting if the info/service are using their local LAN, this forces the provider for maintain low latency across the network. Thus each element’s latency needs to be measure in microseconds
- Load Balancing - high capacity metro network combined with data centers and multiple bottle necks require a device that will manage the load in the edge of the cloud , going into the cloud, going to the metro network or going across multiple cloud
- Caching and resource optimization – this is a critical element to ensure that data is stored in a way that provides a faster storage system or a storage system closer to the usage of the data, cloud gateway can communicate this information with the cloud and optimize the traffic
To summarize, Cloud Gateway is the gateway connecting the cloud, core and metro Ethernet in the different layers of communication, transport, control, service and ensure assure end-to-end SLA.
Image may be NSFW.
Clik here to view.
We’ve lately released the T-Metro 8000 cloud gateway and service aggregation platform, click here to get more details or contact us to discuss.