Abstract: Systems and methods for supporting dual-port virtual router in a high performance computing environment. In accordance with an embodiment, a dual port router abstraction can provide a simple way for enabling subnet-to-subnet router functionality to be defined based on a switch hardware implementation. A virtual dual-port router can logically be connected outside a corresponding switch port. This virtual dual-port router can provide an InfiniBand specification compliant view to a standard management entity, such as a Subnet Manager. In accordance with an embodiment, a dual-ported router model implies that different subnets can be connected in a way where each subnet fully controls the forwarding of packets as well as address mappings in the ingress path to the subnet.

Abstract: Systems and methods for InfiniBand fabric optimizations to minimize SA access and startup failover times. A system can comprise one or more microprocessors, a first subnet, the first subnet comprising a plurality of switches, a plurality of host channel adapters, a plurality of hosts, and a subnet manager, the subnet manager running on one of the one or more switches and the plurality of host channel adapters. The subnet manager can be configured to determine that the plurality of hosts and the plurality of switches support a same set of capabilities. On such determination, the subnet manager can configure an SMA flag, the flag indicating that a condition can be set for each of the host channel adapter ports.

Abstract: System and method for representing performance management agent (PMA) attributes as subnet management agent (SMA) attributes in a high performance computing environment. In general legacy systems, performance metrics, as collected by a performance management agent, can only be collected via the use of general data packets traversing a subnet. In accordance with an embodiment, by configuring a SMA attribute to comprise the same performance metrics as collected by a PMA, a subnet manager, or another management component, can access the performance metrics in-band via the use of a subnet management packet.

Abstract: Systems and methods to use all incoming multicast (MC) packets as a basis for global unique identifier (GUID) to local identifier (LID) cache contents in a high performance computing environment, in accordance with an embodiment. Since all multicast packets have a Global Route Header (GRH), there is always both a source GID and a source LID defined for an incoming multicast packet. This implies that it is, in general, possible for an HCA implementation to gather information about GID and GUID to LID mappings for any sender node based on all incoming MC packets.

Abstract: Systems and methods for InfiniBand fabric optimizations to minimize SA access and startup failover times. A system can comprise one or more microprocessors, a first subnet, the first subnet comprising a plurality of switches, a plurality of host channel adapters, a plurality of hosts, and a subnet manager, the subnet manager running on one of the one or more switches and the plurality of host channel adapters. The subnet manager can be configured to determine that the plurality of hosts and the plurality of switches support a same set of capabilities. On such determination, the subnet manager can configure an SMA flag, the flag indicating that a condition can be set for each of the host channel adapter ports.

Abstract: System and method for supporting scalable representation of switch port status in a high performance computing environment. In accordance with an embodiment, a scalable representation of switch port status can be provided. By adding a scalable representation of switch port status at each switch (both physical and virtual)—instead of getting all switch port changes individually, the scalable representation of switch port status can combine a number of ports that can scale by just using a few bits of information for each port's status.

Abstract: Systems and methods for supporting coordinated link up handling following a switch reset in a high performance computing environment. Systems and methods can ensure that when a switch of a fabric is rebooted, HCA ports connected to that switch will be set in Active state at the same time even though link training times for different ports may vary with up to several seconds.

Abstract: Systems and methods for InfiniBand fabric optimizations to minimize SA access and startup failover times. Based on the combination of path record parameters defined with values from the Homogeneous Fabric/Subnet flag (or configuration parameter), as well as the GID and SLID information defined by IPoIB ARP requests and responses, there is no additional need for SA requests to obtain path records or path related information in general.

Abstract: A system and method can support a middleware machine environment. The middleware machine environment can include a subnet manager and a plurality of fabric components, wherein each said fabric component is associated with a subnet management agent (SMA). The subnet manager maintains a topology graph that indicates how ports are connected in the subnet. The subnet manager operates to periodically reevaluate one or more states associated with each fabric component in the subnet by determining whether each SMA is responding correctly to a request.

Abstract: System and method for supporting scalable representation of switch port status in a high performance computing environment. In accordance with an embodiment, a scalable representation of switch port status can be provided. By adding a scalable representation of switch port status at each switch (both physical and virtual)—instead of getting all switch port changes individually, the scalable representation of switch port status can combine a number of ports that can scale by just using a few bits of information for each port's status.

Abstract: Systems and methods for supporting coordinated link up handling following a switch reset in a high performance computing environment. Systems and methods can ensure that when a switch of a fabric is rebooted, HCA ports connected to that switch will be set in Active state at the same time even though link training times for different ports may vary with up to several seconds.

Abstract: Systems and methods for providing multicast group multicast local identifier (MLID) dynamic discovery on received multicast messages for a relevant multicast global identifier (MGID) in a high performance computing environment. By allowing InfiniBand (IB) clients to associate local queue pairs (QPs) with the MGID(s) of relevant multicast group(s) without requiring any join request to the subnet manager (SM)/subnet administration (SA), it is possible to receive relevant multicast (MC) messages without imposing the SM/SA overhead of a conventional multicast group join request. After receiving, at an end-node of the subnet, a multicast packet including a multicast global identifier and a multicast local identifier, the end-node can inspect the multicast packet to learn the multicast local identifier and include the learned multicast local identifier in the multicast group record at the end-node for the received multicast global identifier.

Abstract: Systems and methods for providing dual multicast local identifiers (MLIDs) per multicast group to facilitate both full and limited partition members in a high performance computing environment. In accordance with an embodiment, in order to avoid the need for the above special handling of P_Key access violations, as well as to ensure complete isolation between limited partition members in terms of multicast traffic, two MLIDs can be allocated to a single MCG, in accordance with an embodiment. A first MLID can be allocated and used by end-ports for sending from full partition members to both full and limited partition members. Additionally, a second MLID can be allocated and used by end-ports for sending from limited partition members to full partition members. Using this scheme, a limited partition member can avoid sending multicast packets to other limited partition members in the MCG.

Abstract: Systems and methods for path record handling in a fabric without host stack cooperation in a high performance computing environment. In a case where the subnet manager has determined “homogenous subnet/fabric” or “semi-homogenous subnet/fabric” status for the current topology, but is still receiving path queries, the subnet manager can use the relevant status to avoid any route evaluation and generate path record either only based on the configuration status of the requesting port in the homogenous case, or by comparing the configuration status of both ports in the semi-homogenous case.

Abstract: System and method for providing an InfiniBand network device having a vendor-specific attribute that contains a signature of the vendor. An exemplary embodiment can provide an InfiniBand compatible network device from a vendor. The vendor can further define a subnet management attribute, where the subnet management attribute has an attribute identifier in the range of attribute identifiers reserved in the InfiniBand specification. The vendor can define a signature bit string that is a fixed number of bits in length and a number of offset bits, where the number of offset bits indicates the start of the signature bit string within the attribute. The subnet management attribute can be configured with the signature bit string starting where the offset bits indicate, and the subnet management attribute can be included as an attribute of the InfiniBand compatible network device.

Abstract: System and method for supporting a flexible framework for extendable SMA attributes in a high performance computing environment. In accordance with an embodiment, an information attribute can provide for enhancements in a number of areas. For example, in addition to indicating which version of an interface a queried node supports, the information attribute can additionally provide a mask indicating which vendor specific SMA attributes the node supports. In this way, a subnet manager can identify a version of an interface at each node in a subnet, as well as each node's SMA attribute capabilities. In turn, this allows nodes to run different versions of an interface within a same subnet, without introducing confusion.

Abstract: System and method for supporting scalable representation of link stability and availability in a high performance computing environment. A method can provide at attribute at each node in a subnet, wherein the attribute provides a single location at each node for a subnet manager to query the stability and availability of each link connected to the queried node. The attribute can be populated and maintained by a subnet management agent residing at the node.

Abstract: System and method for providing an InfiniBand network device having a vendor-specific attribute that contains a signature of the vendor. An exemplary embodiment can provide an InfiniBand compatible network device from a vendor. The vendor can further define a subnet management attribute, where the subnet management attribute has an attribute identifier in the range of attribute identifiers reserved in the InfiniBand specification. The vendor can define a signature bit string that is a fixed number of bits in length and a number of offset bits, where the number of offset bits indicates the start of the signature bit string within the attribute. The subnet management attribute can be configured with the signature bit string starting where the offset bits indicate, and the subnet management attribute can be included as an attribute of the InfiniBand compatible network device.

Abstract: System and method for supporting scalable representation of link stability and availability in a high performance computing environment. A method can provide at attribute at each node in a subnet, wherein the attribute provides a single location at each node for a subnet manager to query the stability and availability of each link connected to the queried node. The attribute can be populated and maintained by a subnet management agent residing at the node.

Abstract: Systems and methods for using queue pair 1 (QP1) for receiving multicast based announcements in multiple partitions in a high performance computing. In accordance with an embodiment, by extending the scope of QP1 to also include receiving and sending multicast packets in any partition defined for the port, it is possible to implement generic MC based announcement and discovery without requiring the complexity of unique QPs for individual partitions, nor any update of QP configuration as a consequence of change of partition membership.