Abstract: Reinstating access to a system of an admin whose certificate is invalid or expired is disclosed. When the admin's certificate is expired, the admin may send a request for reinstatement to tenant admins. One of the tenant admins, if satisfied as to the admin's identity, can invoke a voting operation that allows the tenant admins to vote on whether to reinstate the admin. If the vote is successful, one of the tenant admins is given temporary privileges or permissions to install the admin's new certificate, after which the admin is reinstated and has access to the system.

Abstract: Preventing, mitigating, and reversing actions in a computing system. A voting mechanism is provided that is configured to block malicious or accidental configuration changes or other actions in a computing system. Risky actions cannot be performed as the voting mechanism requires certain actions to be subject to a vote. The actions are then, based on the voting by other administrators, allowed, disallowed, prevented, or reversed. Further different classes of administrators can participate in the voting operations.

Abstract: Methods and systems for seamlessly changing over between inference models are disclosed. The inference models may be distributed across multiple data processing systems. Provide a seamless changeover, updated inference models and original inference models may be managed in accordance with an update framework. The update framework may ensure that the original inference model continues to operate until all of the portions of the updated inference model are in place and ready to operate. During the update process, the update framework may ensure that redundancy goals continue to be met so that failures of some of the data processing systems are not be fatal to continued operation of at least one of the inference models, such as the original or updated inference model.

Abstract: Methods and systems for managing distribution of inference models throughout a distributed system are disclosed. To manage distribution of inference models, a system may include a data aggregator and one or more data collectors. The data aggregator may obtain a threshold, the threshold indicating an acceptable inference error rate for an inference model. The data aggregator may obtain an inference model based on the threshold by training an inference model, performing a lookup in an inference model lookup table, or via other methods. The data aggregator may optimize the inference model to determine the minimum quantity of computing resources consumed by an inference model in order to generate inferences accurate within the threshold. In order to do so, the data aggregator may simulate the operation of more computationally-costly inference models and less computationally-costly inference models.

Abstract: Compressing files is disclosed. An input file to be compressed is first aligned. Aligning the file includes splitting the file into sequences that can be aligned. The result is a compression matrix, where each row of the matrix corresponds to part of the file. The compression matrix may also serve as a warm start if additional compression is desired. Compression may be performed in stages, where an initial compression matrix is generated in a first stage using larger letter sizes for alignment and then a second compression stage is performed using smaller letter sizes. A consensus sequence id determined from the compression matrix. Using the consensus sequence, pointer pairs are generated. Each pointer pair identifies a subsequence of the consensus matrix. The compressed file includes the pointer pairs and the consensus sequence.

Abstract: One example method includes defining an airgap control policy that specifies a threshold data value, generating a value for a set of data, determining whether the value that has been generated for the data meets or exceeds the threshold data value, and opening the air gap when the value that has been generated for the data meets or exceeds the threshold data value. The airgap is closed automatically when the value that has been generated for the data meets or exceeds the threshold data value.

Abstract: In one or more embodiments, one or more systems, one or more methods, and/or one or more processes may execute a process; provide input data to the process as the process executes; receive output data from the process as the process executes; after executing the process, determine a neural network based at least on the input data to the process and the output data from the process; determine multiple binary neural networks from the neural network; determine a network of multiple logic gates based at least on the multiple binary neural networks of the neural network; and configure an integrated circuit based at least on the network of the multiple logic gates. For example, the integrated circuit may include at least one of a field programmable gate array, an application specific integrated circuit, a programmable array logic, and a complex logic device.

Abstract: Embodiments for automatically optimizing and persisting database views by receiving queries made to a database, wherein each query generates a respective database view, and generating a set of database maintained views generated by the queries. The system obtains, for each generated view, certain telemetry information about a respective view including latency, memory space utilization, and processor utilization, among other factors. It then scores each view of the generated views based on an base score modified by the obtained information to determine which one or more of the generated views to make persistent, and maintains the one or more persistent views to produce an optimized persistent set of database views. It further adapts later queries to use the optimized persistent views.

Abstract: Disk drive reallocation or replacement is disclosed. When performing a data protection operation, a health score is determined for each of the disk drives associated with the data protection operation. Replacement is recommended for each of the disk drives with an unfavorable health score. The recommendation may also identify a drive class based in part on the write or wear pattern. This allows unhealthy drives to be replaced prior to performing the data protection operation.

Abstract: A method includes searching a group of PITs, identifying one of the PITs as having an indicator of an occurrence of an event involving data associated with the identified PIT, restoring the data, running a production system copy using the data, and while the production system copy is running, taking increasingly granular backups of the data until the event is located. The event may be a corruption of the data, or other problem.

Abstract: One example method includes receiving, at a remote site from a production site, copies of production site assets, storing, at the remote site, the copies of the production site assets, using, at the remote site, the copies of the production site assets to restore a temporary production site, running the temporary production site at the remote site, and restoring, from the remote site to the production site, the copies of the production site assets.

Abstract: One method includes receiving a request to restore data to a particular point in time, scanning a snapshot that corresponds to the point in time, based on the scanning, identifying any invalid data segments pointed to by the snapshot, for each of the invalid data segments, identifying a most recent, valid, version of that segment, and based on the request, restoring a set of valid data segments. In this method, no invalid segments are restored.

Abstract: A method includes accessing a group that comprises a group of PITs, replaying the PITs according to respective times at which the snapshots were taken, analyzing the PITs as they are being replayed, and based on the analyzing, identifying an event that has occurred within a time frame spanned collectively by the PITs. Replaying the PITs includes presenting the PITs, in order from oldest to newest, as a continuous stream of events.

Abstract: A system can determine a first output from an explainable artificial intelligence risk model based on a first input, wherein the first input indicates a first computing configuration, and wherein the first output indicates a first predicted maintenance cost of the first computing configuration during a time period. The system can determine a second output from the explainable artificial intelligence risk model based on a second input, wherein the second input indicates a second computing configuration that differs from the first computing configuration, and wherein the second output indicates a second predicted maintenance cost of the second computing configuration during the time period. The system can, in response to determining that the second predicted maintenance cost is less than the first predicted maintenance cost, saving an indication of a difference between the second predicted maintenance cost and the first predicted maintenance cost.

Abstract: Opportunistically transmitting backups through a time-limited air gap. A data protection system may predict rates of changes for one or more applications. The predicted rate of change allows a size of corresponding backups to be estimated. If there is time during which an air gap is available (closed), at least of the backups is selected and opportunistically transmitted to a vault through the air gap.

Abstract: A system can train an artificial intelligence risk model to produce a trained model, wherein labeled training data for the training comprises respective features of user accounts and products, and corresponding labels of respective support costs applicable to supporting the products. The system can perform reconstructive self-supervised learning on a group of features of a user account to produce a complete group of features that are specified for the user account. The system can, in response to applying an input to the trained model, wherein the input comprises the complete group of features and a product of the products, produce an output that indicates a predicted cost that corresponds to the input.

Abstract: Redistributing disks based on disk wear patterns is disclosed. The wear patterns of disk drives in a storage system are learned or determined. When a restore operation is performed, the volumes to disk drive mappings are changed to balance the overall wear pattern of the storage system. This insures that, after the restore operation, disks that had comparatively lower wear levels are used more heavily while disks that had comparatively higher wear levels are used less heavily.

Abstract: A forensic kit with a granular infected backup. A forensic engine may evaluate a production system that is infected with malware or other corruption and generate a forensic kit. The forensic kit may include copies of components of the production system that are infected or that are sufficiently related to infected components. The forensic kit may be provided to investigators.

Abstract: A system can fit an artificial intelligence risk model to data based on labeled training data to produce a fitted model, wherein the labeled training data comprises respective features of users and products, and corresponding labels of respective maintenance costs applicable to the products, and wherein the fitted model comprises a tree model that is configured to differentiate between groups of the data with differing maintenance cost distributions. The system can, in response to applying a first input to the fitted model, produce an output that indicates a predicted maintenance cost distribution, wherein the first input comprises a feature of a user of the users and a product of the products.

Abstract: Methods and systems for anomaly detection in a distributed system are disclosed. To manage anomaly detection, a system may include an anomaly detector and one or more data collectors. The anomaly detector may detect anomalies in data obtained from one or more of the data collectors using an inference model. The inference model may be an autoencoder trained to reconstruct data that is intended to match input data to an extent considered acceptable by the system. To accurately perform anomaly detection, the inference model may require re-training. Data collected from the one or more data collectors may be used to re-train the inference model as needed. Following anomaly detection and/or inference model re-training, the data may be discarded to remove the data from the anomaly detector.