Tungsten Fabric有哪些特点？如何运作？如何部署？如何编排？……这是很多初入TF社区的技术朋友的困惑。
为统一答疑，我们将整理出Tungsten Fabric架构有关的内容，陆续在公众号发布，感谢关注，也可以后台留言给我们，把你最希望了解的问题告诉我们，小编去帮你寻找答案。As virtualization becomes the key technology to provide public and private cloud services, virtualization technology (for example, using a network of L2 VMware, as well as the OpenStack Nova, Neutron or ML2 network) so far widely used, the network scale is becoming increasingly evident. Tungsten Fabric provides a highly scalable virtual network platform, designed to support multi-tenant environments largest network, while supporting multiple coordinator.
Since few data center deployments are truly "green field", so almost always required to be deployed on the new infrastructure workloads integrated with the deployment before the work load and network. This article describes a set of deployment scenarios, where the deployment of a new cloud infrastructure, and also needs to coexist with the existing infrastructure.
Tungsten Fabric common use case:
• achieve highly scalable and flexible platform as a service and software as a service in the data center OpenStack managed
virtual network • Use Kubernetes container management system, including Hat OpenShift Red
• VMware vCenter allowed to run new or existing of Tungsten Fabric virtualization environment using the virtual network between the virtual machines
• the gateway router connected to the virtual network Tungsten Fabric physical network coverage and use BGP peering network directly through the underlying network data center.
These use cases can be deployed in any combination to meet the specific requirements of various deployment scenarios.
Tungsten Fabric main features are as follows:
The key feature is support for major use cases:
• Virtual network uses encapsulation tunnel between virtual hosts
• for virtual machines and containers coordinator of the open-source plug-ins
• Based on application-based security policy tags
• Integration with VMware stack of business processes
• using BGP, SNAT, and the underlying network to an external network
Because both use the same controller and forwarding components in each implementation, thus Tungsten Fabric provides a consistent interface to manage all of its connected environment supported, and can in different coordinator (either virtual machines or containers) between provides seamless connection management workload, and the external network to the destination.
Tungsten Fabric main features
Tungsten Fabric using OpenStack and Kubernetes coordinator management and implementation of virtual network in a cloud environment. Use overlay network between vRouters Tungsten Fabric running on each host. It is based on proven and standard network technology, now supported by the world's leading provider of wide area network, but re-virtualized workloads for data center and cloud automation, ranging from large corporate data centers to smaller telecommunications companies POPs. It provides a number of enhancements on the network implementation coordinator of the unit, including:
• Highly scalable multi-tenant network
• Multi-tenant IP address management
• DHCP, ARP proxy, in order to avoid flooding the network
• broadcast and multicast traffic edge efficient replication
• local, each tenant DNS resolution
distributed firewall with access control lists •
• policy-based security applications
• distributed load balancing across hosts
• network address Translation (1: 1 floating IP and distributed SNAT )
• use service links virtual networking features
• IPv4 and IPv6 dual stack support
• BGP peer routers and gateways
• BGP as a service (BGPaaS), for the distribution network routing between customer and service provider networks in the private management
Focus on micro letter: TF Chinese community
-mail: [email protected]