Abstract: Techniques described herein relate to a method for deploying workflows with data management services. The method may include identifying a service update event; identifying a service sub-tree based on a service call graph; generating an update sequence for the service sub-tree; predicting an update window for the service sub-tree; selecting a first service of the service sub-tree based on the update sequence, wherein the first service includes a first standby service instance and a first active service instance; generating a backup of a first portion of a services shared data volume repository associated with the first service; applying an update to the first standby service instance to obtain a first updated active service instance; making a first determination that a first performance and reliability check associated with the first updated active service instance is below a threshold; applying the update to a second standby service instance.

Abstract: A method for managing a plurality of storage devices includes obtaining, by a storage device cleaning manager, a set of self-monitored statistics, performing an initial concern analysis to generate an initial concern prediction for each of the plurality of storage devices in a storage system, wherein the set of self-monitored statistics are associated with the plurality of storage devices, updating a cleaning policy based on the initial concern prediction, obtaining input/output (I/O) statistics, after updating the cleaning policy based on the initial concern prediction, performing a secondary concern analysis using the I/O statistics to generate a secondary concern prediction for each of the plurality of storage devices, wherein the I/O statistics are associated with the plurality of storage devices, further updating the cleaning policy, and performing a cleaning of at least a portion of the plurality of storage devices based on the updated cleaning policy.

Abstract: Embodiments described herein relate to techniques for placing backup data based on health scores. The techniques may include: obtaining data items associated with a first data domain restorer; obtaining data items associated with a second data domain restorer; making a prediction that the first data domain restorer is operating normally; making a prediction that the second data domain restorer is operating normally; assigning a confidence value to the first prediction; making a classification of the first data domain restorer in a first group based on the confidence value; assigning a confidence value to the second prediction; making a classification of the second data domain restorer in a second group based on the confidence value; and performing a data backup to the first data domain restorer from a first computing device based on the classification and a first service level required for the first computing device.

Abstract: Non-volatile Random Access Memory (NVR) on a storage system may be dynamically converted between use as temporary memory in a memory context and use as persistent memory in a storage context. NVR (e.g., embodied as DIMM) may be utilized in a hybrid capacity, where some of the NVR is used as memory and some of the NVR is used as storage, and where NVR memory is converted to memory as needed, dynamically as I/O is being processed using the NVR. A host system may be directly connected to an internal switching fabric of the data storage system without an intervening component of the storage system (e.g., a director) controlling access of the host system to the internal fabric or to the memory. The host system may provision and use the NVR as storage by directly communicating with the NVR over the internal fabric, for example, using RDMA.

Abstract: Methods and systems support managed use of a Storage Class Memory (SCM) by one or more applications operating on an IHS (Information Handling System). The operations that are supported by an IHS processor for flushing data from the SCM are determined. Applications are identified that operate using the persistent data storage capabilities of the SCM. The SCM flushing operations invoked by each these applications are monitored. The utilization of the SCM by each of the first plurality of applications is determined based at least in part on the monitored flushing operations by each application. The utilization of the SCM may also be based on calculated metrics of SCM utilization by the respective applications. The applications are classified based their determined SCM utilizations. Based on the classifications of SCM utilization, a subset of the applications may be identified for removal from use of the SCM.

Abstract: A method and system for policy class based data migration. Specifically, the method and system disclosed herein entail dynamically changing policy classes with which any given data migration may be associated while the given data migration is transpiring. In transitioning between policy classes, different levels of resources, available to different policy classes, respectively, may be allocated to supporting the given data migration.

Abstract: From among physical storage devices (PSDs) of a storage system, a set of two or more of the PSDs that are eligible for scrubbing may be determined; and from among the set, a relative eligibility of the PSDs may be determined. Conformance prediction analysis may be applied to determine the set and the relative eligibility of PSDs of the set. The conformance prediction analysis may determine a scrubbing eligibility classification (e.g., label), and a confidence value for the classification, which may serve as the relative eligibility of the PSD. The eligible PSDs may be ranked in an order according to determined confidence values, and may be further classified according to such order. The future workload of the storage system may be forecasted, and the scrubbing of PSDs may be scheduled based on the forecasted workload of the system and the relative eligibilities of the set of PSDs.

Abstract: Techniques described herein relate to a method for forecasting backup failures. Such techniques may include: obtaining data items associated with backup jobs; writing entries in a time series database, the entries comprising successful backup jobs and failed backup jobs; performing a first analysis to predict future failed backup jobs based on the entries in the time series database to obtain a future backup job failure predictions; performing a second analysis to determine a confidence prediction for each of the future backup job failure predictions; ranking the future backup job failure predictions based on the second analysis; performing a third analysis to determine at least one variable leading to each of the future backup job failure predictions; and sending results of the second analysis and the third analysis to an administrator of a data domain.

Abstract: Embodiments are directed to a method and system of forecasting a disk drive survival period in a data storage network, by obtaining operating system data and manufacturer data for the disk drive to create a dataset, screening the dataset to identify a number of features to be selected for model creation, wherein the data set includes censored data and non-censored data, and performing, in an analytics engine, semi-parametric survival analysis on the data set using transfer learning on the model to provide a time-based failure prediction of the disk drive. A graphical user interface provides to a user the failure prediction in one of text form or graphical form.

Abstract: Techniques described herein relate to a method for forecasting backup failures. Such techniques may include: obtaining data items associated with backup jobs; writing entries in a time series database, the entries comprising successful backup jobs and failed backup jobs; performing a first analysis to predict future failed backup jobs based on the entries in the time series database to obtain a future backup job failure predictions; performing a second analysis to determine a confidence prediction for each of the future backup job failure predictions; ranking the future backup job failure predictions based on the second analysis; performing a third analysis to determine at least one variable leading to each of the future backup job failure predictions; and sending results of the second analysis and the third analysis to an administrator of a data domain.

Abstract: Embodiments described herein relate to techniques for placing backup data based on health scores. The techniques may include: obtaining data items associated with a first data domain restorer; obtaining data items associated with a second data domain restorer; making a prediction that the first data domain restorer is operating normally; making a prediction that the second data domain restorer is operating normally; assigning a confidence value to the first prediction; making a classification of the first data domain restorer in a first group based on the confidence value; assigning a confidence value to the second prediction; making a classification of the second data domain restorer in a second group based on the confidence value; and performing a data backup to the first data domain restorer from a first computing device based on the classification and a first service level required for the first computing device.

Abstract: Embodiments described herein relate to a method for probabilistically forecasting the state of hardware components. The method may include obtaining data items corresponding to a hardware component and performing an analysis of the hardware component. The analysis may include making a variety of probability predictions as to whether a label from among a set of possible labels is likely to be the correct label. The set of probabilities from the aforementioned analysis are then analyzed to determine which predicted label has the tightest range, and the prediction with the tightest range for a certain label is displayed to a user in a ranked fashion that includes a quantity of such probability prediction ranges. Such a display may allow an administrator to take action as to which hardware components should be replaced and in what order.

Abstract: Systems and methods are provided for context-aware maintenance window identification. In some embodiments, an Information Handling System (IHS) may include a processor and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution, cause the IHS to: estimate a completion time of a maintenance operation; predict future usage of the IHS; identify a time window for the maintenance operation based upon the estimation and the prediction.

Abstract: Embodiments are described for an autonomously and dynamically allocating resources in a distributed network based on forecasted a-priori CPU resource utilization, rather than a manual throttle setting. A multivariate (CPU idle %, disk I/O, network and memory) rather than single variable approach for Probabilistic Weighted Fuzzy Time Series (PWFTS) is used for forecasting compute resources. The dynamic throttling is combined with an adaptive compute change rate detection and correction. A single spike detection and removal mechanism is used to prevent the application of too many frequent throttling changes. Such a method can be implemented for several use cases including, but not limited to: cloud data migration, replication to a storage server, system upgrades, bandwidth throttling in storage networks, and garbage collection.

Abstract: A first cloud vendor is registered by a backup application. A file is moved from backup storage of the backup application to the first cloud vendor. A reference is maintained at the backup storage to the first file residing at the first cloud vendor. A second cloud vendor is registered by the backup application. The backup application directs a migration of the file from the first cloud vendor to the second cloud vendor without recalling the file to the backup storage. A reference maintained at the backup storage is updated to refer to the file now residing at the second cloud vendor. The updated reference is maintained at the backup storage.

Abstract: A method and system for dynamic backup policy handshaking. Specifically, the method and system disclosed herein entail optimizing storage space utilization for backup, archiving, and/or disaster recovery-purposed data storage. That is, based on time projections until the data storage reaches capacity, the utilization of the remaining storage space may be optimized without compromising data protection in order to prolong the use of the data storage. In prolonging the utilization of the data storage, tiered data backup policies may be adjusted.

Abstract: A file is backed up to a backup system. A first cloud storage provided by a first cloud provider is connected to the backup system and the file is moved to the first cloud storage. Metadata is created at the backup system to reference the file moved to the first cloud storage. A second cloud storage is connected to the backup system. The file is moved from the first cloud storage to the second cloud storage. The metadata at the backup system is updated to reference the file now residing at the second cloud storage.

Abstract: Embodiments are described for an autonomously and dynamically allocating resources in a distributed network based on forecasted a-priori CPU resource utilization, rather than a manual throttle setting. A multivariate (CPU idle %, disk I/O, network and memory) rather than single variable approach for Probabilistic Weighted Fuzzy Time Series (PWFTS) is used for forecasting compute resources. The dynamic throttling is combined with an adaptive compute change rate detection and correction. A single spike detection and removal mechanism is used to prevent the application of too many frequent throttling changes. Such a method can be implemented for several use cases including, but not limited to: cloud data migration, replication to a storage server, system upgrades, bandwidth throttling in storage networks, and garbage collection.

Abstract: Systems, devices, and methods for reducing the impact of persistent storage failures. Specifically, a system may monitor persistent storages and generate predictions of when such storages are likely to fail. The generated predictions may be used to proactively address potential future failures of the persistent storages. A failure prediction system may generate predictions of future persistent storage failures in a manner that is computationally efficient. To generate the predictions, the system may utilize at least two prediction frameworks (e.g., trained machine learning models). The first of the prediction frameworks may generate accurate predictions at a higher computational cost than the second prediction framework. The second predictions framework may be a refined version of the first prediction framework that generates predictions in a computationally efficient manner. The second prediction framework may utilize smaller amounts of data for generating predictions than the first prediction framework.

Abstract: A file is backed up to a backup system. A first cloud storage provided by a first cloud provider is connected to the backup system and the file is moved to the first cloud storage. Metadata is created at the backup system to reference the file moved to the first cloud storage. A second cloud storage is connected to the backup system. The file is moved from the first cloud storage to the second cloud storage. The metadata at the backup system is updated to reference the file now residing at the second cloud storage.