Abstract: In one set of embodiments, a computer system can receive an input array including n elements and sort the input array to generate a sorted array including the n elements in ascending or descending element order. The computer system can further convert the sorted array into a tree-ordered array including the n elements in a tree order, where the tree order corresponds to a descending order of nodes in a binary tree created from the n elements. The computer system can then place the first m<n elements of the tree-ordered array in first memory tier of the computer system and place the remaining n?m elements of the tree ordered array in a second memory tier of the computer system, where the second memory tier has a slower access time than the first memory tier.

Abstract: Disclosed are various embodiments for decreasing the amount of time spent processing interrupts by switching contexts in parallel with processing an interrupt. An interrupt request can be received during execution of a process in a less privileged user mode. Then, the current state of the process can be saved. Next, a switch from the less privileged mode to a more privileged mode can be made. The interrupt request is then processed while in the more privileged mode. Subsequently or in parallel, and possibly prior to completion of the processing the interrupt request, another switch from the more privileged mode to the less privileged mode can be made.

Abstract: Disclosed are various embodiments for high throughput reclamation of pages in memory. A first plurality of pages in a memory of the computing device are identified to reclaim. In addition, a second plurality of pages in the memory of the computing device are identified to reclaim. The first plurality of pages are prepared for storage on a swap device of the computing device. Then, a write request is submitted to a swap device to store the first plurality of pages. After submission of the write request, the second plurality of pages are prepared for storage on the swap device while the swap device completes the write request.

Abstract: Disclosed are various embodiments for optimized memory tiering. An ideal tier size for a first memory and an ideal tier size for a second memory can be determined for a process. Then, a host computing device can be identified that can accommodate the ideal tier size for the first memory and the second memory. Subsequently, the process can be assigned to the host computing device.

Abstract: Disclosed are various embodiments for optimized memory tiering. A first page can be allocated in a first memory for a process, the first memory being associated with a first memory tier. Accesses of the first page by the process during execution of the process can be monitored. Then, accesses of the first page by the process during execution of the process can be compared to an allocation policy to make a first determination to move the contents of the first page from the first memory to a second memory associated with a second memory tier. Next, the contents of the first page can be copied from the first memory to a second page in the second memory in response to the first determination.

Abstract: Disclosed are various embodiments for decreasing the amount of time spent processing interrupts by switching contexts in parallel with processing an interrupt. An interrupt request can be received during execution of a process in a less privileged user mode. Then, the current state of the process can be saved. Next, a switch from the less privileged mode to a more privileged mode can be made. The interrupt request is then processed while in the more privileged mode. Subsequently or in parallel, and possibly prior to completion of the processing the interrupt request, another switch from the more privileged mode to the less privileged mode can be made.

Abstract: Disclosed are various embodiments for high throughput reclamation of pages in memory. A first plurality of pages in a memory of the computing device are identified to reclaim. In addition, a second plurality of pages in the memory of the computing device are identified to reclaim. The first plurality of pages are prepared for storage on a swap device of the computing device. Then, a write request is submitted to a swap device to store the first plurality of pages. After submission of the write request, the second plurality of pages are prepared for storage on the swap device while the swap device completes the write request.

Abstract: Disclosed are various approaches for decreasing the latency involved in reading pages from swap devices. These approaches can include setting a first queue in the plurality of queues as a highest priority queue and a second queue in the plurality of queues as a low priority queue. Then, an input/output (I/O) request for an address in memory can be received.

Abstract: The present techniques prioritize transmission of feature information associated with data copies in data transmission, and selectively transmit the data copies when necessary. As the size of the feature information is small compared with the data copies, and as feature information is transmitted using prioritized messages, delays caused by congestion in the network are reduced. Based on the use of prioritized messages, the present techniques provide solutions for replicating data across data centers when conducting various data operations including reading, writing, and modifying data. The present techniques also replicate data across data centers while providing consistency, access locality, and disaster tolerance.

Abstract: The present techniques prioritize transmission of feature information associated with data copies in data transmission, and selectively transmit the data copies when necessary. As the size of the feature information is small compared with the data copies, and as feature information is transmitted using prioritized messages, delays caused by congestion in the network are reduced. Based on the use of prioritized messages, the present techniques provide solutions for replicating data across data centers when conducting various data operations including reading, writing, and modifying data. The present techniques also replicate data across data centers while providing consistency, access locality, and disaster tolerance.

Abstract: The present techniques prioritize transmission of feature information associated with data copies in data transmission, and selectively transmit the data copies when necessary. As the size of the feature information is small compared with the data copies, and as feature information is transmitted using prioritized messages, delays caused by congestion in the network are reduced. Based on the use of prioritized messages, the present techniques provide solutions for replicating data across data centers when conducting various data operations including reading, writing, and modifying data. The present techniques also replicate data across data centers while providing consistency, access locality, and disaster tolerance.

Abstract: An embodiment of a method of writing erasure coded data swaps a new data block for an old data block within a stripe of erasure coded data. The stripe of erasure coded data comprises data blocks and a redundancy block. The method computes a redundancy update parameter for the redundancy block using the new data block and the old data block. The method updates the redundancy block using the redundancy block, the redundancy update parameter, and a mathematical operator. An embodiment of a method of recovering erasure coded data obtains a lock on a subset of data blocks and one or more redundancy blocks. The method determines whether the subset of the data blocks and the one or more redundancy blocks includes sufficient blocks to restore the stripe. If not, the method relaxes the lock to allow at least one update of the one or more redundancy blocks while not allowing any writes of the data blocks. The method uses the subset to restore the stripe.

Abstract: The present techniques prioritize transmission of feature information associated with data copies in data transmission, and selectively transmit the data copies when necessary. As the size of the feature information is small compared with the data copies, and as feature information is transmitted using prioritized messages, delays caused by congestion in the network are reduced. Based on the use of prioritized messages, the present techniques provide solutions for replicating data across data centers when conducting various data operations including reading, writing, and modifying data. The present techniques also replicate data across data centers while providing consistency, access locality, and disaster tolerance.

Abstract: An embodiment of a method of seeking consensus among computer processes begins with a first step of saving a new timestamp in a timestamp array for a particular process. The method continues with a second step of determining whether a most recent entry in a decision array includes a previously established consensus decision. In a third step, if the most recent entry does not include the previously established consensus decision, the method saves a proposed decision as a consensus decision. Otherwise, in a fourth step, the method saves the previously established consensus decision as the consensus decision. In a fifth step, if a most recent timestamp in the timestamp array continues to be the new timestamp, the method returns the consensus decision. Otherwise, in a sixth step, the method returns an abort indicator.

Abstract: Method and apparatus for determining causal relations between a plurality of intercommunicating nodes. In various embodiments, trace data is input that describe inter-node communication. The trace data may include for each message sent between nodes a timestamp that indicates a time at which the message was sent, a source identifier that identifies a node from which the message was sent, and a destination identifier that identifies a node to which the message was sent. For each of one or more nodes, a determination may be made as to whether one or more causal relations exist between a first set of messages destined to the node and a second set of messages sourced from the node and destined to at least one other node. A causal relation may exist as a function of a probability distribution of delay values that are differences between timestamps of messages in the second set and timestamps of messages in the first set. From the nodes and causal relations a processor-readable representation is generated.

Abstract: A method for calculating a probability distribution that represents probabilities of occurrence of time differences between causally related incoming messages and outgoing messages of a node in a system of communicating nodes is described. The method comprises obtaining a set of time differences between times at which incoming messages arrive at a node and times at which outgoing messages leave the node, the set of time differences including time differences between causally related incoming messages and outgoing messages and time differences between incoming messages and outgoing messages that are not causally related. The method also comprises processing the set of time differences using a statistical estimation algorithm to calculate the probability distribution that represents probabilities of occurrence of time differences between causally related incoming messages and outgoing messages. An apparatus configured to execute the methods is also described.