Abstract: Techniques for identifying and remedying performance issues of Virtualized Network Functions (VNFs) are discussed. An example system includes processor(s) configured to: process VNF Performance Measurement (PM) data received from a network Element Manager (EM) for a VNF; determine whether the VNF has a negative performance issue based on the VNF PM data; request that the EM create a Virtualization Resource (VR) PM job associated with a VR of the VNF when the VNF has the negative performance issue; process VR PM data received from the EM; determine whether to restart the VNF based on the VR PM data and the VNF PM data; and request a network function virtualization orchestrator (NFVO) to restart the VNF based on a determination to restart the VR.

Abstract: The method relates to a method for checking a correct operation of an application running in a cloud environment, the method comprising at a health checking entity, monitoring a reception of health reports generated by the application, each health report comprising at least information allowing an operation status of the application to be determined, determining whether an anomaly is detected in the reception of the health reports, wherein, when an anomaly is detected in the reception, triggering a transmission of a new health report to be transmitted by the application, determining whether the application is operating correctly after triggering the transmission of the new health report, wherein, if the application is not operating correctly after triggering the transmission of the new health report, informing a control entity of the application that the application is not operating correctly.

Abstract: Techniques in advanced deep learning provide improvements in one or more of cost, accuracy, performance, and energy efficiency. The deep learning accelerator is implemented at least in part via wafer-scale integration. The wafer comprises a plurality of processor elements, each augmented with redundancy-enabling couplings. The redundancy-enabling couplings enable using redundant ones of the processor elements to replace defective ones of the processor elements. Defect information gathered at wafer test and/or in-situ, such as in a datacenter, is used to determine configuration information for the redundancy-enabling couplings.

Abstract: Disclosed herein are systems and method for disaster recovery using application streaming. In one aspect, a method includes generating a backup image of a computing system having at least one installed application and user data. The generating involves including the user data in the backup image and actively excluding program data files of the at least one installed application from the backup image. The method includes determining an application package specifying the installed application. The application package is stored at an application streaming service. Responsive to detecting a disaster recovery event, the method further includes copying the user data from the backup image to a recovery computing instance, and executing, on the recovery computing instance, a remote application from the application streaming service based on the determined application package.

Abstract: A distributed system can include a server outside of a clinical environment and a connectivity adapter and a plurality of infusion pumps within the clinical environment. The connectivity adapter can monitor microservices that measure the quality of connectivity adapter's performance. If the performance is below a threshold level, a message indicating poor performance can be sent to the server. The message can be sent when a prior message relating to poor performance has not already been sent within a predetermined time period.

Abstract: A computing device receives, from a remote device, one or more abnormality detection models for detecting anomaly of one or a plurality of applications with respect to one or more resource usage, obtains resource usage data associated with each application, determines, using the one or more abnormality detection models, whether each application has anomaly in a resource usage, and takes an action in response. The abnormality detection model is built using a machine learning technique at the remote device based on data collected from multiple devices. The resource usage data includes information about applications used by the computing device, e.g., usage or energy consumption of one or more hardware components or one or more services accessed by an application on the device. An abnormality detection result may include likelihoods of an application falling into an abnormality level in a plurality of abnormality levels.

Abstract: An aspect includes generating a data result and a special case indicator based on an instruction and at least one input data operand. Outputting the data result to a processor core. Outputting the first condition code to the processor core prior to outputting the data result to the processor core. Generating a second condition code based on the data result and the special case indicator. Performing a check by comparing the first condition code and the second condition code and flagging an error to the processor core upon the first condition code being different from the second condition code.

Abstract: A test system includes a controller that creates test jobs to simulate a load for testing a target system or service. The jobs specify test data from a data store. The controller adds the test jobs to a job queue at a rate. Workers access jobs from the job queue and request data from the data store as specified in each job and send the request data to the service. Approximations (e.g., metric values indicating performance of the worker host resources, or performance of the system under test or performance of resources that implement the system under test, other feedback, etc.) of actual system characteristics such as the actual number of workers are obtained. Based on the metric values, the controller modulates various characteristics of the load generation system, such as a rate at which jobs are added to the job queue or the number of workers processing jobs.

Abstract: A system, method, and computer-readable medium for performing a system failure repair operation, comprising: receiving information regarding symptoms related to a faulty device; storing the information with other historical information regarding the symptoms; receiving additional information as the faulty device is diagnosed; indicating whether a repair recommendation is provided for the faulty device; and using the stored information, historical information, and additional information to provide a repair recommendation if indicating shows no repair recommendation.

Abstract: Examples of systems described herein include a virtualized file servers. Examples of virtualized file servers described herein may support disaster recovery of the virtualized file server. Accordingly, examples of virtualized file servers may support metadata fixing procedures to update metadata in a recovery setting. Examples of virtualized file servers may support hypervisor-agnostic disaster recovery.

Abstract: Anomaly detection for one or more streams of time-series data can use an encoder/decoder pair, such as in a variational autoencoder (VAE) in combination with an aggregator or classifier, such as a random isolation forest (RIF). A particular application relates to detecting anomalies in network updates in a large number of network devices that can transmit the updates to a collector for analysis. The encoder/decoder pair can include a neural network with long short-term memory cells or similar type cells. Using the combination, a single anomaly score can be produced from multiple streams of the time-series data.

Abstract: A computer system having a host in communication with a data storage device that is coupled to the host via a peripheral bus and a host interface. The data storage device has a controller, non-volatile storage media; and firmware containing instructions to configure the operations of the controller. The computer system allows a user of the system to specify a service option for operating the data storage device beyond its end of normal service life. The user selected option is stored in the computer system and controls the operations of the storage device beyond its end of service life.

Abstract: The present disclosure relates to systems, methods, and computer readable media for diagnosing and mitigating memory impact events, such as memory leaks, high memory usage, or other memory issues causing a host node from performing as expected on a cloud computing system. The systems described herein involve receiving locally generated memory usage data from a plurality of host nodes. The systems described herein may aggregate the memory usage data and determine a memory impact diagnosis based on a subset of the aggregated memory usage data. The systems described herein may further apply a mitigation model for mitigating the memory impact event. The systems described herein provide an end-to-end solution for diagnosing and mitigating a variety of memory issues using a dynamic and scalable system that reduces a negative impact of memory leaks and other memory issues on a cloud computing system.

Abstract: A method, computer program product, computing system, and system for using hot-plug virtual disks in granular recovery are described. The method may include receiving, at a cloud computing site, a backup configuration including a list of protected resources associated with a cloud computing tenant. The method may further include mounting a hot-plug virtual disk to a virtual machine including the protected resources associated with the cloud computing tenant. The method may also include collecting metadata corresponding to the protected resources associated with the cloud computing tenant. The method may additionally include generating, at the cloud computing site, a backup of the virtual machine including the protected resources associated with the cloud computing tenant.

Abstract: Systems and methods for detecting and managing incidents are disclosed. In one embodiment, a method for detecting an incident includes receiving issue data created for an issue tracking system; analyzing the received issue data over a predetermined interval; determining whether a potential incident has occurred based on the analysis; upon determining that a potential incident has occurred, creating an incident management assistant program; identifying one or more relevant users to communicate an alert to; and communicating the alert to the identified relevant users, the alert including a pointer to the incident management program.

Abstract: Embodiments of the invention are directed to techniques for detecting anomalous values in data streams using forecasting models. In some embodiments, a computer can receive a value of a data stream comprising a plurality of data values, where the received value corresponds to a time interval and previously received values each correspond to a previous time interval. Models can be selected based on the time interval, where each of the models has a different periodicity. For each of the selected models, the computer may generate a score by generating a prediction value based on the model and generating the score based on the prediction value and the received value. A final score can then be generated based on the scores. Next, a score threshold can be generated. If the final score exceeds the score threshold, the computer may generate a notification that indicates that the data value is an anomaly.

Abstract: According to examples, an apparatus may include a first processing stage device to process data, a second processing stage device to process the processed data, and a control device. The control device may access first performance information of the first processing stage device, determine, from the accessed first performance information, an operational state of the first processing stage device, and determine whether a countermeasure is to be taken based on the determined operational state of the first processing stage device, in which the countermeasure is to improve the operational state of the first processing stage device. The control device may also, based on a determination that a countermeasure is to be taken, output an instruction for the countermeasure to be implemented.

Abstract: A system includes software and a computing device configured to test the software in parallel. The computing device may obtain a first test case from a plurality of test cases. The first test case may be configured to cause the software to access a subset of data entries in a database. A plurality of locks facilitates exclusive access to the plurality of data entries. The computing device may further attempt to acquire a subset of the locks that correspond to the subset of the data entries. The computing device may further determine when to execute the first test case. Doing so involves either: (i) failing to acquire all of the subset of the locks and delaying execution of the first test case, or (ii) successfully acquiring all of the subset of the locks and executing the first test case in parallel with the second test case.

Abstract: Techniques for managing a storage disk involve monitoring a duration of a fault of a faulted storage disk, wherein the faulted storage disk includes a first disk slice configured to store metadata and a second disk slice configured to store user data. The techniques further involve, in response to the duration reaching a first threshold value, replacing the first disk slice with a first available disk slice in a first non-faulted storage disk. The techniques further involve, in response to the duration reaching a second threshold value greater than the first threshold value, replacing the second disk slice with a second available disk slice in a second non-faulted storage disk. Accordingly, fault monitoring windows with different lengths are applied to disk slices for different logical tiers in the faulted storage disk. In this way, the reliability of data of a metadata tier can be effectively improved.

Abstract: A system stores logs representing events that occur in the system based on executable instructions executed by the system, for example, by processes executing within the system or by applications. The system analyzes the logs to determine the root cause of the error or event that resulted in generation of the log. The system clusters logs to determine clusters of logs. The system analyzes logs of each cluster to determine a root cause of errors resulting in logs belonging to the cluster. For any new error log that is received, the system determines the cluster to which the error log belongs and takes action based on the root cause associated with the cluster, for example, sending an alert message or performing automatic remediation.