Abstract: Technologies for allocating resources across data centers include a compute device to obtain resource utilization data indicative of a utilization of resources for a managed node to execute a workload. The compute device is also to determine whether a set of resources presently available to the managed node in a data center in which the compute device is located satisfies the resource utilization data. Additionally, the compute device is to allocate, in response to a determination that the set of resources presently available to the managed node does not satisfy the resource utilization data, a supplemental set of resources to the managed node. The supplemental set of resources are located in an off-premises data center that is different from the data center in which the compute device is located. Other embodiments are also described and claimed.

Abstract: Technologies for load balancing a storage network include a system. The system includes circuitry to adjust routing rules in a network interface controller to deliver a packet from one of multiple uplinks to one of any physical functions, circuitry to remap, in response to a failure of a switch, a port from one physical function to another physical function, and circuitry to communicate control data between a software defined network controller and one or more agents in one or more host endpoints with a hierarchical distributed hashing table.

Abstract: Systems, apparatuses and methods may provide for technology that detects a first failure in a first storage server, wherein the first storage server is connected to a first non-volatile memory (NVM) via a switch, selects a second storage server that is connected to the first NVM via the switch, wherein the first storage server and the second storage server are in a storage cluster, and configures the second storage server to host first data resident on the first NVM, wherein configuring the second storage server to host the first data bypasses a cluster-wide rebalance of the storage cluster.

Abstract: Technologies for providing advanced management of power usage limits in a disaggregated architecture include a compute device. The compute device includes circuitry configured to execute operations associated with a workload in a disaggregated system. The circuitry is also configured to determine whether a present power usage of the compute device is within a predefined range of a power usage limit assigned to the compute device. Additionally, the circuitry is configured to send, to a device in the disaggregated system and in response to a determination that the present power usage of the present compute device is not within the predefined range of the power usage limit assigned to the present compute device, offer data indicative of an offer to reduce the power usage limit assigned to the present compute device to enable a second power utilization limit of another compute device in the disaggregated system to be increased.

Abstract: Technologies for fast distributed storage recovery include a distributed storage system that includes multiple controller nodes and multiple target nodes. Each controller node is coupled to a corresponding target node via a storage fabric. Each target node stores replica data. The system identifies a failed node and a corresponding node that was coupled to the failed node. If the failed node is a controller node, the corresponding node is a target node. If the failed node is a target node, the corresponding node is a controller node. The system instantiates a replacement node, adds the replacement node to the system, and couples the replacement node to the corresponding node. The system may direct a backup target node to copy replica data to the replacement target node via the storage fabric. Other embodiments are described and claimed.

Abstract: Technologies for allocating resources across data centers include a compute device to obtain resource utilization data indicative of a utilization of resources for a managed node to execute a workload. The compute device is also to determine whether a set of resources presently available to the managed node in a data center in which the compute device is located satisfies the resource utilization data. Additionally, the compute device is to allocate, in response to a determination that the set of resources presently available to the managed node does not satisfy the resource utilization data, a supplemental set of resources to the managed node. The supplemental set of resources are located in an off-premises data center that is different from the data center in which the compute device is located. Other embodiments are also described and claimed.

Abstract: Technologies for monitoring node cluster health include a plurality of managed nodes of anode cluster communicatively coupled across a data network to a resource manager server. The resource manager server is configured to receive health data, via an out-of-band network, from each of the managed nodes of the node cluster. The resource manager server is further configured to identify whether a managed node of the plurality of managed nodes has indicated a failure, determine a cause of the failure, and classify the failure as being one of a soft failure or a hard failure as a function of the received health data and the cause of the failure. Additionally, the resource manager server is configured to transmit a health state change event to each of the other managed nodes of the plurality of managed nodes of the node cluster. Other embodiments are described herein.

Abstract: Technologies for managing burst bandwidth requirements are disclosed. In the illustrative embodiment, a software-defined network (SDN) controller monitors storage devices in a data center. If a storage device fails, the SDN controller manages the bandwidth used to replicate the data that was stored on the failed storage device. The SDN controller may allocate an initial amount of bandwidth based on one or more parameters of the storage device, and the SDN controller may increase the bandwidth in a series of discrete steps. In another embodiment, the SDN controller may predict a bandwidth burst based on sequential writes at a storage sled from several compute devices, and allocate bandwidth accordingly in a tiered manner.

Abstract: Technologies for load balancing a storage network include a system. The system includes circuitry to adjust routing rules in a network interface controller to deliver a packet from one of multiple uplinks to one of any physical functions, circuitry to remap, in response to a failure of a switch, a port from one physical function to another physical function, and circuitry to communicate control data between a software defined network controller and one or more agents in one or more host endpoints with a hierarchical distributed hashing table.

Abstract: A method performed by a networking switch in an object storage system. The method includes receiving a first packet from a network comprising an object ID and a data object. The method includes generating a replica for the data object. The method includes generating an object ID for the replica of the data object. The method includes determining a destination storage node for the replica of the data object. The method includes sending a second packet from the networking switch to the destination storage node. The second packet includes the object ID for the replica of the data object and the replica of the data object.

Abstract: Technologies for accelerated memory lookups include a computing device having a processor and a hardware accelerator. The processor programs the accelerator with a search value, a start pointer, one or more predetermined offsets, and a record length. Each offset may be associated with a pointer type or a value type. The accelerator initializes a memory location at the start pointer and increments the memory location by the offset. The accelerator may read a pointer value from an offset, set the memory location to the pointer value, and repeat for additional offsets. The accelerator may read a value from the offset and compare the value to the search value. If the values match, the accelerator returns the address of the matching value to the processor. If the values do not match, the accelerator searches a next record based on the record length. Other embodiments are described and claimed.

Abstract: Technologies for load balancing a storage network include a system. The system includes circuitry to adjust routing rules in a network interface controller to deliver a packet from one of multiple uplinks to one of any physical functions, circuitry to remap, in response to a failure of a switch, a port from one physical function to another physical function, and circuitry to communicate control data between a software defined network controller and one or more agents in one or more host endpoints with a hierarchical distributed hashing table.

Abstract: Technologies for providing advanced management of power usage limits in a disaggregated architecture include a compute device. The compute device includes circuitry configured to execute operations associated with a workload in a disaggregated system. The circuitry is also configured to determine whether a present power usage of the compute device is within a predefined range of a power usage limit assigned to the compute device. Additionally, the circuitry is configured to send, to a device in the disaggregated system and in response to a determination that the present power usage of the present compute device is not within the predefined range of the power usage limit assigned to the present compute device, offer data indicative of an offer to reduce the power usage limit assigned to the present compute device to enable a second power utilization limit of another compute device in the disaggregated system to be increased.

Abstract: An object node apparatus, system, and method are described. An apparatus can include a lockless-mode controller configured to communicatively couple to a plurality of storage resources and to a plurality of processor cores each preassigned to process a specific type of sub-task at a different preassigned storage resource that is configured to receive object input/output (I/O) only from the preassigned core, the lockless-mode controller being further configured to receive a plurality of object I/O messages from one or more clients, each to perform an object I/O task, divide each object I/O task into a plurality of sub-tasks, identify a specific sub-task type for each sub-task, and send each sub-task for each specific sub-task type to a processor core preassigned to process the specific sub-task type, wherein the sub-tasks include storage operations related to storing sub-object data in, or retrieving sub-object data from, the preassigned storage resource for each processor core.

Abstract: Technologies for storage cluster quality of service (QoS) management include multiple compute devices in communication via a storage network. A controller node monitors network traffic of the storage cluster. The network traffic includes a replication traffic class and a rebuild traffic class. The controller node determines whether burst bandwidth is required by the storage cluster and, if so, applies a group policy indicative of burst bandwidth to the storage cluster. The group policy may be applied to an end to end path of the storage cluster. Applying the group policy may include setting one or more bits or fields of an overlay network header of network traffic of the storage cluster. Other embodiments are described and claimed.

Abstract: Technologies for providing efficient message polling include a compute device. The compute device includes circuitry to determine a memory location to monitor for a change indicative of a message from a device connected to a local bus of the compute device. The circuitry is also to determine whether data at the memory location satisfies reference data. Additionally, the circuitry is to process, in response to a determination that the data at the memory location satisfies the reference data, one or more messages in a message queue associated with the memory location.

Abstract: Technologies for managing burst bandwidth requirements are disclosed. In the illustrative embodiment, a software-defined network (SDN) controller monitors storage devices in a data center. If a storage device fails, the SDN controller manages the bandwidth used to replicate the data that was stored on the failed storage device. The SDN controller may allocate an initial amount of bandwidth based on one or more parameters of the storage device, and the SDN controller may increase the bandwidth in a series of discrete steps. In another embodiment, the SDN controller may predict a bandwidth burst based on sequential writes at a storage sled from several compute devices, and allocate bandwidth accordingly in a tiered manner.

Abstract: Technologies for fast distributed storage recovery include a distributed storage system that includes multiple controller nodes and multiple target nodes. Each controller node is coupled to a corresponding target node via a storage fabric. Each target node stores replica data. The system identifies a failed node and a corresponding node that was coupled to the failed node. If the failed node is a controller node, the corresponding node is a target node. If the failed node is a target node, the corresponding node is a controller node. The system instantiates a replacement node, adds the replacement node to the system, and couples the replacement node to the corresponding node. The system may direct a backup target node to copy replica data to the replacement target node via the storage fabric. Other embodiments are described and claimed.

Abstract: Technologies for automatic workload detection and cache quality of service (QoS) policy determination include a computing device that executes a workload. The computing device receives a data item associated with the workload, such as a file, block, or page. The computing device extracts a workload feature vector from the data item and determines a workload grouping based on the workload feature vector. The computing device determines a cache QoS policy based on the workload grouping. The cache QoS policy may be determined based on predetermined priority levels associated with workload groupings or with a machine learning model. The computing device applies the cache QoS policy to the workload. The cache QoS policy may be a guaranteed or maximum bandwidth, guaranteed or maximum I/O operation rate, maximum latency, caching mode, cache space allocation, or other cache QoS policy. Other embodiments are described and claimed.

Abstract: Examples are disclosed for determining or using server transaction latency information. In some examples, a network input/output device coupled to a server may be capable of time stamping information related to ingress request and egress response packets for a transaction. For these examples, elements of the server may be capable of determining transaction latency values based on the time stamped information. The determined transaction latency values may be used to monitor or manage operating characteristics of the server to include an amount of power provided to the server or an ability of the server to support one or more virtual servers. Other examples are described and claimed.