Abstract:
An example method for simultaneously supporting, on a single VIC adapter (i.e. on a single network interface), RDMA transport according to multiple RoCE versions is disclosed. The method includes providing the VIC adapter with an indication of a RoCE version used for a particular RDMA connection between two compute nodes identified by a certain QP identifier, and then automatically configuring Egress (EG) packet classifier and flow table of the VIC adapter to encapsulate outgoing RDMA commands and automatically configuring Ingress (IG) packet classifier and flow table of the VIC adapter to decapsulate incoming RoCE packets for that RDMA connection according to the indicated RoCE version. Automatic encapsulation and decapsulation according to any one of the multiple RoCE versions that may be used for a particular RDMA connection, i.e. on a per-QP basis, advantageously allows co-existence of routable (i.e. RoCEv2) and non-routable (i.e. RoCEv1) RDMA solutions on a single network interface.
Abstract:
An example method for facilitating policy-driven storage in a microserver computing environment is provided and includes receiving, at an input/output (I/O) adapter in a microserver chassis having a plurality of compute nodes and a shared storage resource, policy contexts prescribing storage access parameters of respective compute nodes and enforcing the respective policy contexts on I/O operations by the compute nodes, in which respect a particular I/O operation by any compute node is not executed if the respective policy context does not allow the particular I/O operation. The method further includes allocating tokens to command descriptors associated with I/O operations for accessing the shared storage resource, identifying a violation of any policy context of any compute node based on availability of the tokens, and throttling I/O operations by other compute nodes until the violation disappears.
Abstract:
An example method for transformation of Peripheral Component Interconnect Express (PCIe) compliant virtual devices in a server in a network environment is provided and includes receiving, during runtime of the server, a request to change a first configuration of a PCIe compliant virtual device to a different second configuration, identifying a bridge on a PCIe topology below which the virtual device is located, issuing a simulated secondary bus reset to the bridge, the virtual device being reconfigured according to the change in configuration after the simulated secondary bus reset is issued, re-enumerating below the bridge after the change in configuration completes without rebooting the server, and updating the PCI topology with the virtual device in the second configuration. A virtual interface card adapter traps the simulated secondary bus reset, removes the virtual device from the PCI topology, and reconfigures the virtual device from the first configuration to the second configuration.
Abstract:
A method is provided in one example and includes creating a staging queue in a virtual interface card (VIC) adapter firmware of a server based on a log policy; receiving a log message from a VIC driver in the server; copying the log message to the staging queue; generating a VIC control message comprising the log message from the staging queue; and sending the VIC control message to a switch.
Abstract:
An example method for facilitating policy-driven storage in a microserver computing environment is provided and includes receiving, at an input/output (I/O) adapter in a microserver chassis having a plurality of compute nodes and a shared storage resource, policy contexts prescribing storage access parameters of respective compute nodes and enforcing the respective policy contexts on I/O operations by the compute nodes, in which respect a particular I/O operation by any compute node is not executed if the respective policy context does not allow the particular I/O operation. The method further includes allocating tokens to command descriptors associated with I/O operations for accessing the shared storage resource, identifying a violation of any policy context of any compute node based on availability of the tokens, and throttling I/O operations by other compute nodes until the violation disappears.
Abstract:
A method is provided in one example embodiment and includes receiving by a network element a request from a network device connected to the network element to update a shared resource maintained by the network element; subsequent to the receipt, identifying a Base Address Register Resource Table (“BRT”) element assigned to a Peripheral Component Interconnect (“PCI”) adapter of the network element associated with the network device, wherein the BRT points to the shared resource; changing an attribute of the identified BRT from read-only to read/write to enable the identified BRT to be written by the network device; and notifying the network device that the attribute of the identified BRT has been changed, thereby enabling the network device to update the shared resource via a Base Address Register (“BAR”) comprising the identified BRT.
Abstract:
An example method for adaptively coalescing remote direct memory access (RDMA) acknowledgements is provided. The method includes determining one or more input/output (I/O) characteristics of RDMA packets of a plurality of queue pairs (QPs) on a per-QP basis, each QP identifying a respective RDMA connection between a respective first compute node and a respective second compute node. The method further includes determining an acknowledgement frequency for providing acknowledgements of the RDMA packets on a per-QP basis (i.e., a respective acknowledgement frequency is set for each QP) based on the determined one or more I/O characteristics for each QP.
Abstract:
An example method for flexible remote direct memory access resource configuration in a network environment is provided and includes determining whether sufficient remote direct memory access (RDMA) resources are available in a network environment to satisfy a request for RDMA resources, inserting the requested RDMA resources into a network profile, associating the network profile with a network interface endpoint in the network, and communicating the network profile over the network to a virtual interface card (VIC) adapter that processes RDMA packets, the VIC adapter configuring the requested RDMA resources in the VIC adapter's hardware and the requested RDMA resources being mapped to a host memory for use by the network interface endpoint. In specific embodiments, the VIC adapter allocates and identifies a region in local memory for managing the requested RDMA resources and reserved for the network interface endpoint.
Abstract:
An example method for facilitating low latency remote direct memory access (RDMA) for microservers is provided and includes generating queue pair (QPs) in a memory of an input/output (I/O) adapter of a microserver chassis having a plurality of compute nodes executing thereon, the QPs being associated with a remote direct memory access (RDMA) connection between a first compute node and a second compute node in the microserver chassis, setting a flag in the QPs to indicate that the RDMA connection is local to the microserver chassis, and performing a loopback of RDMA packets within the I/O adapter from one memory region in the I/O adapter associated with the first compute node of the RDMA connection to another memory region in the I/O adapter associated with the second compute node of the RDMA connection.
Abstract:
A method is provided in one example and includes creating a staging queue in a virtual interface card (VIC) adapter firmware of a server based on a log policy; receiving a log message from a VIC driver in the server; copying the log message to the staging queue; generating a VIC control message comprising the log message from the staging queue; and sending the VIC control message to a switch.