A network system for a data center is described in which a switch fabric provides interconnectivity such that any servers may communicate packet data to any other of the servers using any of a number of parallel data paths. Moreover, according to the techniques described herein, edge-positioned access nodes, permutation devices and core switches of the switch fabric may be configured and arranged in a way such that the parallel data paths provide single L2/L3 hop, full mesh interconnections between any pairwise combination of the access nodes, even in massive data centers having tens of thousands of servers. The access nodes may be arranged within access node groups, and permutation devices may be used within the access node groups to spray packets across the access node groups prior to injection within the switch fabric, thereby increasing the fanout and scalability of the network system.

A network system for a data center is described in which a switch fabric provides interconnectivity such that any servers may communicate packet data to any other of the servers using any of a number of parallel data paths. Moreover, according to the techniques described herein, edge-positioned access nodes, permutation devices and core switches of the switch fabric may be configured and arranged in a way such that the parallel data paths provide single L2/L3 hop, full mesh interconnections between any pairwise combination of the access nodes, even in massive data centers having tens of thousands of servers. The access nodes may be arranged within access node groups, and permutation devices may be used within the access node groups to spray packets across the access node groups prior to injection within the switch fabric, thereby increasing the fanout and scalability of the network system.

Techniques for providing a non-blocking fabric in a network are described. A network controller determines the network requirement for various network traffic types on the network and determines the allocation of resources across the network needed to establish a midlay, including midlay components on the network. The network controller then establishes the midlay on the network according to the determined allocation. At least one of the midlay components is a virtually non-blocking fabric for high-priority traffic or fully non-blocking fabric for deterministic traffic.

Techniques for providing a non-blocking fabric in a network are described. A network controller determines the network requirement for various network traffic types on the network and determines the allocation of resources across the network needed to establish a midlay, including midlay components on the network. The network controller then establishes the midlay on the network according to the determined allocation. At least one of the midlay components is a virtually non-blocking fabric for high-priority traffic or fully non-blocking fabric for deterministic traffic.

A network switch includes a field programmable gate array (FPGA) and a processor. The FPGA is configured to transfer a processing target packet to a transfer destination, based on transfer definition information, and to transfer a copy of the processing target packet to the processor. The processor is configured to delete an entry of the transfer definition information based on a transfer record information, and to update the transfer record information based on the copy of the processing target packet.

A network switch includes a field programmable gate array (FPGA) and a processor. The FPGA is configured to transfer a processing target packet to a transfer destination, based on transfer definition information, and to transfer a copy of the processing target packet to the processor. The processor is configured to delete an entry of the transfer definition information based on a transfer record information, and to update the transfer record information based on the copy of the processing target packet.

Network controllers are described that enable creation of logical interconnects between logical routers of different, isolated virtual networks and for auto-generation and deployment of routing policies to control “leaking” of select routes amongst the different virtual networks. In one example, a network controller includes a memory and processing circuitry configured to identify a source logical router of a first virtual network and a destination logical router of a second virtual network implemented on one or more physical devices of a switch fabric, form a policy defining one or more rules for controlling leaking of one or more of the routes through a logical router interconnect from the source logical router to the destination logical router, and push the policy to the one or more physical devices of the switch fabric for application to communications through the logical router interconnect.

Some embodiments of the invention provide a forwarding element that can be configured through in-band data-plane messages from a remote controller that is a physically separate machine from the forwarding element. The forwarding element of some embodiments has data plane circuits that include several configurable message-processing stages, several storage queues, and a data-plane configurator. A set of one or more message-processing stages of the data plane are configured (1) to process configuration messages received by the data plane from the remote controller and (2) to store the configuration messages in a set of one or more storage queues. The data-plane configurator receives the configuration messages stored in the set of storage queues and configures one or more of the configurable message-processing stages based on configuration data in the configuration messages.

Some embodiments of the invention provide a forwarding element that can be configured through in-band data-plane messages from a remote controller that is a physically separate machine from the forwarding element. The forwarding element of some embodiments has data plane circuits that include several configurable message-processing stages, several storage queues, and a data-plane configurator. A set of one or more message-processing stages of the data plane are configured (1) to process configuration messages received by the data plane from the remote controller and (2) to store the configuration messages in a set of one or more storage queues. The data-plane configurator receives the configuration messages stored in the set of storage queues and configures one or more of the configurable message-processing stages based on configuration data in the configuration messages.

A storage system determines source addresses, and destination addresses in a storage system, for network traffic. The storage system determines a hash algorithm, from a plurality of hash algorithms. The hash algorithm is to be used across the source addresses for load-balancing the network traffic to the destination addresses. The storage system determines that the hash algorithm more closely meets one or more load-balancing criteria than at least one other hash algorithm, of the plurality of hash algorithms. The storage system distributes the network traffic from the source addresses to the destination addresses in the storage system, with load-balancing according to the determined hash algorithm.