H04L2101/645

Server network interface level failover

A technique includes hosting an Internet Storage Name Service (iSNS) server on a network switch to provide access to a plurality of subnets; and configuring the iSNS server to be an active server for a first subnet of a plurality of subnets and to be a backup server for a second subnet of the plurality of subnets.

Fibre channel over ethernet (FCoE) pinning system

A Fibre Channel over Ethernet (FCoE) pinning system includes a first Fibre Channel Forwarder (FCF) device that is coupled to a Fibre Channel (FC) storage device through a first FCF device port, and that is coupled to a first server device through a second FCF device port that is included on an aggregated interface. The FCF device sends an FCF advertisement through the second FCF device port. The FCF device receives, from the first server device at the second FCF device port in response to sending the FCF advertisement through the second FCF device port, a server message that includes a first server port identifier of a first server port. The FCF device associates the first server port identifier with the second FCF device port to form a first FCF group and pins the first FCF group for FCoE traffic.

Link aggregated fibre channel over ethernet system

A link aggregated FCoE system includes a target device, a first FCF device coupled to the target device and a LAG, and a second FCF device coupled to the LAG and to the first FCF device by an ICL. The first and second FCF devices are each associated with a common FCF MAC address. The first FCF device receives, through the LAG, first FCoE data traffic directed to the common FCF MAC address and including a target device destination identifier and, in response, forwards the first FCoE data traffic to the target. The second FCF device receives, through the LAG, second FCoE data traffic directed to the common FCF MAC address and including the target device destination identifier and, in response, forwards the second FCoE data traffic to the first FCF device so that the first FCF device may forward the second FCoE data traffic to the target device.

System and method of assigning admin partition membership based on switch connectivity in a high-performance computing environment

System and method of assigning admin partition membership based on switch connectivity in a high performance computing environment. An exemplary embodiment can provide an admin partition for use in a subnet of a network fabric. The admin partition can be defined by a P_Key, and the P_Key can be stored in an admin partition registry. A subnet manager that executes from a node within the subnet can be provided. The subnet manager can obtain information that a subnet switch is associated with the admin partition. The subnet manager can also receive a list of identifiers (IDs), each identifier corresponding to a host channel adapter that is directly connected to the subnet switch. Each of the IDs from the list can be added to the admin partition registry and a relationship between each of the added IDs and the P_Key can be created in the registry.

Virtual node port based virtual fabric support

Embodiments of the present invention include systems and methods for transmitting data in a fibre channel (FC) network. The system includes a gateway that has: a port at which a plurality of data frames that are transmitted from a plurality of virtual node (VN) ports in a server are received, where data frames from at least two of the plurality of VN ports are related to different virtual fabrics (vfabrics) of a FC network.

VIRTUAL SWITCH FABRICS IN CONVERGED NETWORKS
20200236066 · 2020-07-23 ·

A Fibre Channel over Ethernet (FCoE) network can be inexpensively extended by additional switches (220x) that do not have a full FCoE functionality, e.g. do not have full E-port functionality. The additional switches may or may not have unique Domain IDs. A virtual switch fabric can be extended by such additional switches.

STORAGE SYSTEM WITH NETWORK-WIDE CONFIGURABLE DEVICE NAMES
20200213274 · 2020-07-02 ·

A storage system in one embodiment comprises at least one storage array comprising a plurality of storage devices. The storage array is configured to communicate over one or more networks with a plurality of host devices. The storage array is further configured to receive one or more commands from a given one of the host devices directing that a particular user-defined name be assigned to a particular one of the storage devices, and responsive to receipt of the one or more commands, to determine if any other user-defined name is already assigned to the particular storage device, and if there is no other user-defined name already assigned to the particular storage device, to assign the particular user-defined name to the particular storage device and to store information characterizing the assignment of the particular user-defined name to the particular storage device in persistent memory of the storage array.

SYSTEM AND METHOD FOR APPLICATION OF VIRTUAL HOST CHANNEL ADAPTER CONFIGURATION POLICIES IN A HIGH-PERFORMANCE COMPUTING ENVIRONMENT
20200195494 · 2020-06-18 ·

Systems and methods for application of virtual host channel adapter configuration policies. An exemplary embodiment can provide a host channel adapter (HCA) with a control circuit and a memory, and at least one port operative to connect the HCA to a network fabric. A software driver of the HCA can provide a control API that exposes functionality to hyper-privileged software executing on the host. The hyper privileged software, through the control API, can request registration of a virtual HCA (vHCA) with a virtual function of the HCA. The HCA can forward the registration request to a subnet manger, which can retrieve configuration policies from a cache and return the policies to the HCA for implementation with the virtual function of the HCA.

Computing device and method for performing a fabric deployment in a data center

Method and computing device performing a fabric deployment in a data center. The computing device stores a configuration file comprising first and second IPv6 base prefixes, and a fabric identifier. The computing device generates a host identifier, a fabric-wide IPv6 prefix by combining the first IPv6 base prefix and the fabric identifier, and a fabric-wide IPv6 address by combining the fabric-wide IPv6 prefix and the host identifier. The computing device determines a local node identifier and a local link identifier for a communication interface of the computing device. The computing device performs a (secure) neighbor discovery procedure for determining a remote node identifier and a remote link identifier for a communication interface of a remote computing device. The computing device generates a link IPv6 address based on the second IPv6 base prefix and at least some of: the local node and link identifiers, and the remote node and link identifiers.

Storage system with network-wide configurable device names

A storage system in one embodiment comprises at least one storage array comprising a plurality of storage devices. The storage array is configured to communicate over one or more networks with a plurality of host devices. The storage array is further configured to receive one or more commands from a given one of the host devices directing that a particular user-defined name be assigned to a particular one of the storage devices, and responsive to receipt of the one or more commands, to determine if any other user-defined name is already assigned to the particular storage device, and if there is no other user-defined name already assigned to the particular storage device, to assign the particular user-defined name to the particular storage device and to store information characterizing the assignment of the particular user-defined name to the particular storage device in persistent memory of the storage array.