Abstract: The technology disclosed herein enables an L3 network fabric including one or more spine switches having a leaf-spine topology to be self-expanded. In a particular embodiment, a method provides transferring one or more probe messages from each of the spine switches. The probe messages detect whether new computing nodes have been attached to the communication network. The method further provides receiving a reply to at least one of the probe messages. The reply identifies a new computing node that is not yet included in the L3 fabric.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to improve workload domain management of virtualized server systems. An example apparatus includes a resource status analyzer to determine a health status of a first virtualized server of a workload domain, compare the health status to a decomposition threshold based on a policy, and transfer a workload of the first virtualized server to a second virtualized server of the workload domain when the health status satisfies the decomposition threshold. The example apparatus further includes a resource deallocator to deallocate the first virtualized server from the workload domain to a pool of virtualized servers to execute the workload using the second virtualized server.

Abstract: Methods, apparatus, systems, and articles of manufacture to improve packet flow among virtualized servers are disclosed. An example apparatus includes memory, and hardware to execute instructions to generate a load balance list identifying first ones of virtualized network resources having respective values of a utilization status parameter that satisfy a first threshold, in response to at least a number of the first virtualized network resources not satisfying a load balance list threshold, update the load balance list to additionally identify second ones of the virtualized network resources, the second ones of the virtualized network resources having respective values of the utilization status parameter that satisfy a second threshold, the first threshold different from the second threshold, and adjust a policy of a physical hardware resource corresponding to one or more of the virtualized network resources based on the load balance list.

Abstract: A disclosed example includes determining that first virtual machines (VMs) corresponding to a first percentage resource utilization indicative of resource usage of a first rack cause the first rack to generate a threshold amount of heat; determining, based on a second percentage resource utilization indicative of resource usage of a second rack, to migrate the first VMs to the second rack to reduce a temperature in a first room by at least the threshold amount of heat; migrating the first VMs from the first rack of the first room to the second rack of a second room; placing the first rack in a low-power state based on the migration of the first VMs to the second rack; and sending a temperature control signal to a climate control system to adjust a cooling process of the climate control system based on the first rack being in the low-power state.

Abstract: A synchronizer is to synchronize a first MAC address from a first MAC address table of a first TOR switch to a second MAC address table of a second TOR switch, the first MAC address corresponding to a host, the first NIC to form a first active link between the first host and the first TOR switch, and store a first switch descriptor in association with the first MAC address in the second MAC address table, the first switch descriptor indicative of the first TOR switch. A packet transmitter is to, for a first packet to be delivered to the first host, send the first packet from the second TOR switch to the first TOR switch, without sending the first packet from the second TOR switch to the first host, based on the first switch descriptor stored in association with the first MAC address in the second MAC address table.

Abstract: A synchronizer is to synchronize a first MAC address from a first MAC address table of a first TOR switch to a second MAC address table of a second TOR switch, the first MAC address corresponding to a host, the first NIC to form a first active link between the first host and the first TOR switch, and store a first switch descriptor in association with the first MAC address in the second MAC address table, the first switch descriptor indicative of the first TOR switch. A packet transmitter is to, for a first packet to be delivered to the first host, send the first packet from the second TOR switch to the first TOR switch, without sending the first packet from the second TOR switch to the first host, based on the first switch descriptor stored in association with the first MAC address in the second MAC address table.

Abstract: Methods, apparatus, systems, and articles of manufacture to perform network fabric migration in virtualized servers are disclosed and described. An example apparatus a layer detector to determine a first network fabric layer of a communication network by sending a first probe packet from a first network resource to a second network resource via a communication link and determine to migrate the first network fabric layer to a second network fabric layer based on whether the first network resource receives a reply probe packet from the second network resource in response to the first probe packet.

Abstract: A disclosed example of managing a network includes a packet receiver to receive a packet at a second switch from a first switch via an inter-switch link between the first and second switches; a packet analyzer to determine whether the packet includes a load-based teaming (LBT) egress control value; a packet transmitter to handle the packet according to an LBT policy when the packet analyzer determines the packet does not include the LBT egress control value; and a packet modifier to terminate the packet at the second switch when the packet analyzer determines the packet does include the LBT egress control value.

Abstract: A disclosed example of managing a network include receiving a packet at a first top-of-rack (ToR) switch via a first load-based teaming (LBT) port; determining whether a network location of a destination node of the packet is unknown; based on the network location being unknown, setting an LBT egress control value in the packet, the LBT egress control value to indicate that the packet is not to be transmitted via a second LBT port of a second ToR switch; and sending the packet from the first ToR switch to the second ToR switch via an inter-switch link between the first and second ToR switches and from the first ToR switch to at least a first host that is identified in the packet as the destination node.

Abstract: Methods, apparatus, systems, and articles of manufacture to optimize or otherwise improve packet flow among virtualized servers are disclosed and described. An example apparatus includes a load balance list generator to identify abstracted network resources in a virtualized network based on a utilization status parameter of the abstracted network resources, generate a load balance list including the identified abstracted network resources, the load balance list including a number of the abstracted network resources satisfying a threshold, and generate a network routing configuration for data packets in the virtualized network based on the generated load balance list. The example apparatus further includes a policy adjustor to adjust a policy of a physical hardware resource corresponding to one of the one or more abstracted network resources based on the network routing configuration to distribute a packet flow among the abstracted network resources.

Abstract: An example to preserve packet order in a multi-fabric network includes: a migration detector to, after a first host sends a first packet of a first packet flow via a first active link between a first host and a first top-of-rack (TOR) switch, detect a migration of the first active link to a second active link between the first host and a second TOR switch, the first packet destined to a second host connected to the second TOR switch; and a migration notifier to, in response to the migration, send a migration notification message to cause configuration of the second TOR switch to send a second packet of the first packet flow to the first TOR switch via an inter-switch link between the first and second TOR switches, the second packet destined to the second host.

Abstract: A disclosed example apparatus to manage port allocations for network load balancing includes a telematics network information collector to collect bandwidth utilizations corresponding to physical links of network paths between a source and a destination, the network paths including at least one of a link aggregation group hop or an equal cost multi-path hop; a cost calculator to determine network path costs corresponding to unused protocol ports of a physical server host based on the bandwidth utilizations; a comparator to compare the network path costs; and a port allocator to, based on the comparison, allocate one of the unused protocol ports to an application.

Abstract: A disclosed example of managing a network include receiving a packet at a first top-of-rack (ToR) switch via a first load-based teaming (LBT) port; determining whether a network location of a destination node of the packet is unknown; based on the network location being unknown, setting an LBT egress control value in the packet, the LBT egress control value to indicate that the packet is not to be transmitted via a second LBT port of a second ToR switch; and sending the packet from the first ToR switch to the second ToR switch via an inter-switch link between the first and second ToR switches and from the first ToR switch to at least a first host that is identified in the packet as the destination node.

Abstract: A disclosed example of managing a network include receiving a packet at a first top-of-rack (ToR) switch via a first load-based teaming (LBT) port; determining whether a network location of a destination node of the packet is unknown; based on the network location being unknown, setting an LBT egress control value in the packet, the LBT egress control value to indicate that the packet is not to be transmitted via a second LBT port of a second ToR switch; and sending the packet from the first ToR switch to the second ToR switch via an inter-switch link between the first and second ToR switches and from the first ToR switch to at least a first host that is identified in the packet as the destination node.

Abstract: A disclosed example includes: a temperature predictor to determine a combined ambient air temperature of a data center during a future duration based on 1) a first ambient air temperature corresponding to heat generated by hardware resources in physical server racks when executing workloads and 2) a second ambient air temperature corresponding to a building structure of the data center; a power utilization analyzer to determine a predicted total data center power utilization for the future duration based on a computing power utilization and a climate control power utilization, the climate control power utilization based on a power utilization corresponding to adjusting the combined ambient air temperature to satisfy an ambient air temperature threshold; and a power manager to configure a power supply station to deliver an amount of electrical power during the future duration to satisfy the predicted total data center power utilization.

Abstract: A disclosed example includes determining that first virtual machines (VMs) corresponding to a first percentage resource utilization indicative of resource usage of a first rack cause the first rack to generate a threshold amount of heat; determining, based on a second percentage resource utilization indicative of resource usage of a second rack, to migrate the first VMs to the second rack to reduce a temperature in a first room by at least the threshold amount of heat; migrating the first VMs from the first rack of the first room to the second rack of a second room; placing the first rack in a low-power state based on the migration of the first VMs to the second rack; and sending a temperature control signal to a climate control system to adjust a cooling process of the climate control system based on the first rack being in the low-power state.

Abstract: The disclosure provides an approach for managing an application workload on a computer system that includes data centers. The application workload includes first application instances running on a first data center and second application instances running on a second data center. The method comprises collecting usage data of first application instances, usage data of second application instances, and combining the collected data. The method further comprises evaluating the combined data to determine low health in least one application instance, and restarting the at least one application instance or creating a new application instance. The method further comprises evaluating the combined data to determine whether to change size of the application workload, and in which data center to place a second new application instance if increasing size. The method further comprises contacting a component of the chosen data center to place the second new application instance within a local host machine.

Abstract: A disclosed example of autonomously configuring a virtual network and a physical network in a physical rack includes generating network topologies of hosts based on physical network connection information indicative of physical network connections between the hosts and a top-of-rack switch in the physical rack; determining whether implementing the network topologies of the hosts concurrently in the physical rack is valid based on evaluating the network topologies relative to a network topology validation rule; when implementing the network topologies of the hosts concurrently in the physical rack is valid: configuring a virtual distributed switch in a first one of the hosts based on one of the network topologies; and configuring the top-of-rack switch in communication with the first host based on the one of the network topologies.

Abstract: The disclosure herein describes a virtual extensible local area network (VXLAN) gateway. During operation, the VXLAN gateway receives, from a physical host, an Ethernet packet destined for a virtual machine residing in a remote layer-2 network broadcast domain that is different from a local layer-2 network broadcast domain where the physical host resides. The VXLAN gateway then determines a VXLAN identifier for the received Ethernet packet. The VXLAN gateway further encapsulates the Ethernet packet with the virtual extensible local area network identifier and an Internet Protocol (IP) header, and forwards the encapsulated packet to an IP network, thereby allowing the packet to be transported to the virtual machine via the IP network and allowing the remote layer-2 network broadcast domain and the local layer-2 network broadcast domain to be part of a common layer-2 broadcast domain.

Abstract: Methods, apparatus, and articles of manufacture are disclosed for template driven infrastructure in virtualized server systems. An example apparatus includes a catalog mapper to map a requirement and system information associated with a cloud computing environment including a virtual controller to a first tag including instructions to deploy a first virtual resource corresponding to physical hardware and identify a template including the first tag based on the mapping, a plan generator to generate a plan corresponding to the template including an order of execution of tags including the first tag to deploy virtual resources including the first virtual resource to the cloud computing environment, and a controller invoker to deploy a workload domain to the cloud computing environment to execute an application by executing the instructions associated with the first tag included in the template in response to determining that the virtual controller can execute the plan.