Abstract: Testcase recommendations are generated for a testcase creator application by training a learning function using metadata of previously generated testcases by parsing the metadata into steptasks, and providing the parsed metadata to the learning function to enable the learning function to determine relationships between the steptasks of the previously generated testcases, and using, by the testcase creator application, the trained learning function to obtain a predicted subsequent steptask for a given type of testcase to be generated. Each steptask describes one of the steps of the testcase using a concatenation of a step number of the one of the steps of the testcase, a module and a submodule to be used to perform of the one of the steps of the testcase, and a function to be performed at the one of the steps of the testcase.

Abstract: Aspects of the present disclosure relate to detecting and correcting a storage array's drive sectors to prevent data loss. In embodiments, a storage array's telemetry information is monitored. Further, one or more state or operational metrics of the storage array's storage drives are measured. Additionally, each storage drive is scrubbed based on each drive's relative scrubbing priority defined by the telemetry information and each storage drive's state or operation metrics.

Abstract: A method of processing test execution logs to determine error location and source includes creating a set of training examples based on previously processed test execution logs, clustering the training examples into a set of clusters using an unsupervised learning process, and using training examples of each cluster to train a respective supervised learning process to label data where each generated cluster is used as a class/label to identify the type of errors in the test execution log. The labeled data is then processed by supervised learning processes, specifically a classification algorithm. Once the classification model is built it is used to predict the type of the errors in future/unseen test execution logs. In some embodiments, the unsupervised learning process is a density-based spatial clustering of applications with noise clustering application, and the supervised learning processes are random forest deep neural networks.

Abstract: An amount of storage space required to maintain counter information for a storage system is reduced without reducing a temporal granularity or tracking granularity of the counter information. Rather than periodically recording actual (i.e., raw) counter values for counters, difference (i.e., delta) values may be recorded. For a given counter, a difference (delta value) between a value of the counter for a given point in time (PIT) and a value of the counter for a previous PIT may be determined, and this delta value may be stored as opposed to storing the raw counter value. This delta value may be a significantly smaller value than the raw value. To further reduce the amount of storage space required, no value may be stored for a counter for a given PIT if it is determined that there is no difference between a counter value for the given PIT and a previous PIT.

Abstract: Embodiments of the present disclosure measure a state of a storage group within a storage array. The embodiments also increase or decrease a compression ratio corresponding to input/output (I/O) operations on the storage group based on a target data reduction ratio (DRR) of the storage array, an expected performance envelope, and a compressibility factor of the storage group.

Abstract: Aspects of the present disclosure relate to data deduplication (dedupe). In embodiments, an input/output operation (IO) stream is received by a storage array. In addition, a received IO sequence in the IO stream that matches a previously received IO sequence is identified. Further, a data deduplication (dedupe) technique is performed based on a selected data dedupe policy. The data dedupe policy can be selected based on a comparison of service quality (QoS) related to the received IO sequence and a QoS related to the previously received IO sequence.

Abstract: An aspect of the present disclosure relates to one or more data decryption techniques. In embodiments, an input/output operation (IO) stream including one or more encrypted IOs is received by a storage array. Each encrypted IO is assigned an encryption classification. Further, each encrypted IO is processed based on its assigned encryption classification.

Abstract: An aspect of the present disclosure relates to one or more techniques to identify and resolve storage array errors. In embodiments, an error notification related to a computing device can be received. One or more threads related to the error notification can further be identified. Additionally, an error resolution technique can be performed based on each identified thread.

Abstract: Estimating maintenance for a storage system includes accessing a model that outputs time and materials estimates based on input configuration data, providing configuration data of the storage system to the model, and obtaining an estimate of maintenance time and materials based on the configuration data provided to the model. The model may be provided by a neural network, which may be a self-organized map. Weights of neurons of the self-organized map may be initialized randomly. The model may be initially configured using training data that may include an I/O load of the storage system, memory size of the storage system, a drive count of the storage system, and/or size and parameter information corresponding to hardware being added for the maintenance operation. The training data may include actual time and materials for prior storage system maintenance operations used for the training data. The model may be provided on the storage system.

Abstract: A method of determining feature usage on a set of storage systems deployed across multiple customer sites includes defining metrics related to the features of interest, and pushing the defined metrics to an AIM (Autonomous Infrastructure Module) of an operating system of each storage system. The AIM on each storage system collects data associated with the metrics from the operating system on the storage system. The collected data is aggregated and formatted by the AIM and then used to create an autonomous field telemetry report. Autonomous field telemetry reports are periodically forwarded on a communication network to an analytics engine. The analytics engine parses each autonomous field telemetry report to extract usage information related to the features of interest, loads the parsed data to PostgreSQL staging and historical databases, and uses the parsed data alone or in combination with the historical data to create analytics and visualizations of the analytics.

Abstract: A primary storage system appends a red-hot data indicator to each track of data transmitted on a remote data facility during an initial synchronization state. The red-hot data indicator indicates, on a track-by-track basis, whether the data associated with that track should be stored as compressed or uncompressed data by the backup storage system. The red-hot data indicator may be obtained from the primary storage system's extent-based red-hot data map. If the red-hot data indicator indicates that the track should remain uncompressed, or if the track is locally identified as red-hot data, the backup storage system stores the track as uncompressed data. If the red-hot data indicator indicates that the track should be compressed, the backup storage system compresses the track and stores the track as compressed data. After the initial synchronization process has completed, red-hot data indicators are no longer appended to tracks by the primary storage system.

Abstract: Described herein is a system, and related techniques, for predicting I/O requests that are not necessarily directed to sequential sectors of a physical storage device. In some embodiments, I/O patterns that do not involve sequential-sector access, and that may be relatively long-term patterns, may be recognized. To recognize such patterns, deep machine-learning techniques may be used, for example, using neural networks. Such neural networks may be a recurrent neural network such as, for example, an LSTM-RNN. I/O streams for a workstream may be sampled for specific I/O features to produce a time series of I/O feature values of a workstream, and this time series of data may be fed to a prediction engine, e.g., an LSTM-RNN to predict one or more future I/O features values, and I/O actions may be taken based on these predicted feature values.

Abstract: A method of determining feature usage on a set of storage systems deployed across multiple customer sites includes defining metrics related to the features of interest, and pushing the defined metrics to an AIM (Autonomous Infrastructure Module) of an operating system of each storage system. The AIM on each storage system collects data associated with the metrics from the operating system on the storage system. The collected data is aggregated and formatted by the AIM and then used to create an autonomous field telemetry report. Autonomous field telemetry reports are periodically forwarded on a communication network to an analytics engine. The analytics engine parses each autonomous field telemetry report to extract usage information related to the features of interest, loads the parsed data to PostgreSQL staging and historical databases, and uses the parsed data alone or in combination with the historical data to create analytics and visualizations of the analytics.

Abstract: A method of processing test execution logs to determine error location and source includes creating a set of training examples based on previously processed test execution logs, clustering the training examples into a set of clusters using an unsupervised learning process, and using training examples of each cluster to train a respective supervised learning process to label data where each generated cluster is used as a class/label to identify the type of errors in the test execution log. The labeled data is then processed by supervised learning processes, specifically a classification algorithm. Once the classification model is built it is used to predict the type of the errors in future/unseen test execution logs. In some embodiments, the unsupervised learning process is a density-based spatial clustering of applications with noise clustering application, and the supervised learning processes are random forest deep neural networks.

Abstract: Embodiments of the present disclosure relate to throttling processing threads of a storage device. One or more input/output (I/O) workloads of a storage device can be monitored. One or more resources consumed by each thread of each storage device component to process each operation included in a workload can be analyzed. Based on the analysis, consumption of each resource consumed by each thread can be controlled.

Abstract: A method is used in managing remote replication in storage systems. The method monitors network traffic characteristics of a network. The network enables communication between a first storage system and a second storage system. The method predicts a change in at least one of an application demand of an application of a set of applications executing on the first storage server and a network state of the network, where the set of applications have been identified for performing a replication to the second storage system. Based on the prediction, the method dynamically manages replication of the set of applications in accordance with a performance target associated with each application.

Abstract: Estimating maintenance for a storage system includes accessing a model that outputs time and materials estimates based on input configuration data, providing configuration data of the storage system to the model, and obtaining an estimate of maintenance time and materials based on the configuration data provided to the model. The model may be provided by a neural network, which may be a self-organized map. Weights of neurons of the self-organized map may be initialized randomly. The model may be initially configured using training data that may include an I/O load of the storage system, memory size of the storage system, a drive count of the storage system, and/or size and parameter information corresponding to hardware being added for the maintenance operation. The training data may include actual time and materials for prior storage system maintenance operations used for the training data. The model may be provided on the storage system.

Abstract: A method, computer program product, and computer system for receiving, by a computing device, a first I/O request from a first application in a first storage group assigned to a first storage level. A response time for the first I/O request may be identified. It may be identified that the response time for the first I/O request is outside a pre-determined response time. A delay may be added to a second I/O request received from a second application in a second storage group assigned to a second storage level based upon, at least in part, identifying that the response time for the first I/O request is outside the pre-determined response time.

Abstract: Generative adversarial networks (GAN) are used to model real IO workloads on storage nodes such as storage area networks (SANs) and network-attached storage (NAS). A GAN model is generated in situ on a storage node or in a data center using real traffic, e.g. an IO trace. The GAN model is sent to a modeling system that maintains a repository of GAN models generated from different storage nodes. An IO traffic emulator in the modeling system uses a GAN model to generate a synthetic IO stream that emulates but does not replay a real IO stream. Multiple configurations of test storage nodes may be tested with synthetic IO streams generated from GAN models and the corresponding performance measurements may be stored in a repository and used to generate recommendations, e.g. for storage node configuration to achieve a target performance level based on IO workload.

Abstract: Determining storage of particular data in cache memory of a storage device includes using a first mechanism to determine when to remove the particular data from the cache memory and using a second mechanism, independent from the first mechanism, to inhibit the particular data from being stored in the cache memory independent of whether the first mechanism otherwise causes the particular data to be stored in the cache memory. The first mechanism may remove data from the cache memory that was least recently accessed. The second mechanism may be based, at least in part, on a prediction value of an expected benefit of storing the particular data in the cache memory. The prediction value may be determined based on input data corresponding to measured cache read hits (RH), cache write hits (WH), cache read misses (RM), cache write destage operations (WD), and prefetch reads (PR) for the particular data.