Various implementations described herein relate to systems and methods for predicting and managing drive hazards for Solid State Drive (SSD) devices in a data center, including receiving telemetry data corresponding to SSDs, determining future hazard of one of those SSDs based on an a-priori model or machine learning, and causing migration of data from that SSD to another SSD.

Disclosed herein is a technique to automatically re-bind applications and storage volumes during recovery from planned outage or platform failure and disaster recovery. Such implementations can involve managing volume management information that maps volume claim information with a logical volume identifier associated with persistent volume information; and for receipt of a request for a new volume, referencing the volume management information to retrieve an associated logical volume identifier and the persistent volume information based on the volume claim information associated with the request; and providing the retrieved persistent volume information to a volume provisioning function.

Systems and methods are disclosed that are of retrieving, by a processing device, a codeword stored at a memory sub-system, determining parity data of the codeword, generating additional parity bits based on one or more bits of the parity data of the codeword, and generating host data by decoding the codeword using the additional parity bits.

Efficiently transferring data between tiers in a virtual storage system, including: receiving, by the virtual storage system, a request to write data to the virtual storage system; transforming, within storage provided by a first tier of storage of the virtual storage system, the data to generate transformed data; and migrating, from the first tier of storage to a second tier of storage that is more durable than the first tier of storage of the virtual storage system, at least a portion of the transformed data.

Application-specific prioritization of streaming data replication. Data streamed from connected devices is selectively replicated to data storage clusters based on needs of the applications being served by the data. Data characterization supports prioritized replication processing. Statistical metrics compare streaming data with estimated values to characterize the data for prioritization.

A non-volatile memory express (NVMe) switch is located in between a host and storage. A first storage access command is received from a host via a peripheral computer interface express (PCIe) interface to access the storage. The first storage access command conforms to NVMe and the storage comprises two or more solid-state drives (SSDs). A respective second storage access command is sent to the two or more SSDs based on the first storage access command. A respective completion is received from each of the two or more SSDs based on the respective second storage access command. A completion is sent to the host via the PCIe interface based on the received completions from each of the two or more SSDs.

Embodiments of the present disclosure relate to a memory system, a memory controller, and a method of operating the memory system. According to the embodiments of the present disclosure, when result data obtained by derandomizing data included in a flag area is different from reference data after a random data unit is derandomized based on a seed, it is possible to detect an error occurring in the seed in a process of derandomizing the data and to prevent malfunction of firmware in advance by searching for a target seed and derandomizing the random data unit based on the target seed.

Devices and techniques for an adjustable watchdog in a memory device are disclosed herein. A memory operation command is received at a first time with a memory device from a host. A reset signal is received, with the memory device from the host, at a second time following the first time. A time interval between the first time and the second time is measured. A delay interval for a timer in the memory device to reset the memory device independently of receiving a further reset signal from the host is established based on the measured time interval.

Generating datasets using approximate baselines including receiving, by a source storage system, an instruction to create, on a target storage system, a current snapshot for a source dataset stored on the source storage system, wherein no snapshots for the source dataset exist on the target storage system; selecting, as a baseline dataset, a similar dataset from a plurality of datasets on the source storage system with an existing snapshot on the target storage system, wherein the similar dataset comprises at least a portion of the source dataset; instructing the target storage system to generate a baseline snapshot for the source dataset using a copy of the existing snapshot of the baseline dataset; and transferring, from the source storage system to the target storage system, only a difference between the baseline dataset and the source dataset.

Methods, systems, and devices for read performance techniques for time retention are described. A memory system may store data in a block of memory cells and perform a power cycle operation. Based on performing the power cycle operation, the memory system may determine a first voltage offset associated with the block of memory cells by executing a first read command using an auto-read calibration operation. Based on the first voltage offset, and, in some examples, one or more additional voltage offsets, the memory system may calculate a retention time of data stored in the block of memory cells. The memory system may adjust a read voltage based on the retention time and perform one or more additional read commands.