Abstract: Techniques are provided for incremental snapshot copy to an object store. A list of deallocated block numbers of primary storage of a computing device are identified. Entries for the list of deallocated block numbers are removed from a mapping metafile. A list of changed block numbers corresponding to changes between a current snapshot of the primary storage and a prior copied snapshot copied from the primary storage to the object store is determined. The mapping metafile is evaluated using the list of changed block numbers to identify a deduplicated set of changed block numbers without entries within the mapping metafile. An object, comprising data of the deduplicated set of changed block numbers, is transmitted to the object store for storage as a new copied snapshot.

Abstract: Techniques are provided for implementing a persistent memory storage tier to manage persistent memory of a node. The persistent memory is managed by the persistent memory storage tier at a higher level within a storage operating system storage stack than a level at which a storage file system of the node is managed. The persistent memory storage tier intercepts an operation targeting the storage file system. The persistent memory storage tier retargets the operation from targeting the storage file system to targeting the persistent memory. The operation is transmitted to the persistent memory.

Abstract: Techniques are provided for implementing a unified object format. The unified object format is used to format data in a performance tier (e.g., infrequently accessed data, snapshot data, etc.) into objects that are stored into an object store for low cost, scalable, long term storage compared to storage of the performance tier. With the unified object format, compression of the data may be retained when the data is stored as the objects into the object store. Additional compression may also be provided for the data in the objects. The unified object format includes slot header metadata used to track the location of the data within the object notwithstanding the data being compressed and/or stored at non-fixed boundaries. The slot header metadata may be cached at the performance tier for improved read performance and may be repaired by a repair subsystem (a slot header repair subsystem).

Abstract: Failover methods and systems for a networked storage environment are provided. In one aspect, a read request associated with a first storage object is received, during a replay of entries of a log stored in a non-volatile memory of a second storage node for a failover operation initiated in response to a failure at a first storage node. The second storage node operates as a partner node of the first storage node. The read request is processed using a filtering data structure that is generated from the log prior to the replay and identifies each log entry. The read request is processed when the log does not have an entry associated with the read request, and when the filtering data structure includes an entry associated with the read request, the requested data is located at the non-volatile memory.

Abstract: A system is described. The system includes a processing resource and a non-transitory computer-readable medium, coupled to the processing resource, having stored therein instructions that when executed by the processing resource cause the processing resource to receive a plurality of quality of service (QoS) parameters and client preferences from a client device and manage a QoS policy based on a plurality of QoS objectives included in the received QoS parameters, wherein the plurality of QoS objectives comprise input output operations per second (IOPS), throughput and latency.

Abstract: Systems, methods, and machine-readable media for monitoring a storage system and correcting demand imbalances among nodes in a cluster are disclosed. A performance manager for the storage system may detect performance imbalances that occur over a period of time. When operating below an optimal performance capacity, the manager may cause a volume to be moved from a node with a high load to a node with a lower load to achieve a preventive result. When operating at or near optimal performance capacity, the manager may cause a QOS limit to be imposed to prevent the workload from exceeding the performance capacity, to achieve a proactive result. When operating abnormally, the manager may cause a QOS limit to be imposed to throttle the workload to bring the node back within the optimal performance capacity of the node, to achieve a reactive result. These actions may be performed independently, or in cooperation.

Abstract: Systems and methods for performing a zero-copy volume move between nodes of a distributed storage system are provided. In one example, an approach for performing a zero-copy volume move is proposed in which volume data may be maintained in place within a storage pod and need not be copied to move a given volume between the source node and the destination node. In one embodiment, metadata (e.g., a top-most physical volume block number (PVBN) of a node tree representing the volume at issue) of a write-anywhere file system is copied from the source node to the destination node. Since the storage pod is associated with a global PVBN space that is visible and accessible to all nodes of the distributed storage system, as a result of copying the top-mode PVBN of the volume to the destination node, anything below the top-most PVBN will automatically be visible to the destination node.

Abstract: One or more techniques and/or computing devices are provided for implementing synchronous replication. For example, a synchronous replication relationship may be established between a first storage controller hosting local storage and a second storage controller hosting remote storage (e.g., replication may be specified at a file, logical unit number (LUN), or any other level of granularity). Data operations and offloaded operations may be implemented in parallel upon the local storage and the remote storage. Error handling operations may be implemented upon the local storage and implement in parallel as a best effort on the remote storage, and a reconciliation may be performed to identify any data divergence from the best effort parallel implementation. Storage area network (SAN) operations may be implemented upon the local storage, and upon local completion may be remotely implemented upon the remote storage.

Abstract: Techniques are provided for implementing a garbage collection process and a prediction read ahead mechanism to prefetch keys into memory to improve the efficiency and speed of the garbage collection process. A log structured merge tree is used to store keys of key-value pairs within a key-value store. If a key is no longer referenced by any worker nodes of a distributed storage architecture, then the key can be freed to store other data. Accordingly, garbage collection is performed to identify and free unused keys. The speed and efficiency of garbage collection is improved by dynamically adjusting the amount and rate at which keys are prefetched from disk and cached into faster memory for processing by the garbage collection process.

Abstract: Methods and systems for a networked storage system are provided. One method includes assigning a quality of service (QOS) parameter for a storage volume of a networked storage environment having a first storage node and a second storage node, where the QOS parameter is defined by a throughput value that defines a maximum data transfer rate and a number of input/output (I/O) operations executed within a time period (IOPS); distributing the QOS parameter between the first storage node and the second storage node; determining that throughput credit is available for processing an I/O request for using the storage volume; determining that IOPS credit is available for processing the request by the first storage node; and processing the I/O request when both the throughput credit and the IOPS credit is available.

Abstract: Techniques are provided for lock reservations for shared storage. A reserve command to reserve a storage structure is received by a driver from a node. The reserve command is formatted according to a storage protocol. The driver translates the reserve command into a lease acquire command formatted according to an object store protocol and targeting an object stored within an object store and corresponding to the storage structure. A lease identifier derived from a node identifier of the node is inserted into the lease acquire command. The lease acquire command is routed to the object store for obtaining a lease on the object for granting the node exclusive write access to the object.

Abstract: Systems and methods are described for performing persistent inflight tracking of operations (Ops) within a cross-site storage solution. According to one embodiment, a method comprises maintaining state information regarding a data synchronous replication status for a first storage object of a primary storage cluster and a second storage object of a secondary storage cluster. The method includes performing persistent inflight tracking of I/O operations with a first Op log of the primary storage cluster and a second Op log of the secondary storage cluster, establishing and comparing Op ranges for the first and second Op logs, and determining a relation between the Op range of the first Op log and the Op range of the second Op log to prevent divergence of Ops in the first and second Op logs and to support parallel split of the Ops.

Abstract: Systems and methods are described for using a Deep Reinforcement Learning (DRL) agent to automatically tune Quality of Service (QoS) settings of a distributed storage system (DSS). According to one embodiment, a DRL agent is trained in a simulated environment to select QoS settings (e.g., a value of one or more of a minimum IOPS parameter, a maximum IOPS parameter, and a burst IOPS parameter). The training may involve placing the DRL agent into every feasible state representing combinations of QoS settings, workload conditions, and system metrics for a period of time for multiple iterations, and rewarding the DRL agent for selecting QoS settings that minimize an objective function based on a selected measure of system load. The trained DRL agent may then be deployed to one or more DSSs to constantly update QoS settings so as to minimize the selected measure of system load.

Abstract: Methods and systems for managing storage is provided. One method includes initializing, a storage service user interface (“SSUI”) within a cloud manager user interface (“CMUI”), the SSUI enables a storage service operation selected from a backup operation, a restore operation and a cloning operation associated with a storage object stored at a cloud volume presented by a cloud layer, the CMUI presented by a cloud provider; transmitting, by the SSUI, an authorization token to a Software As a Service (“SaaS”) layer for authenticating deployment of the SSUI; upon validating the authorization token, initializing a SSUI agent to interface with a deployed storage micro-service layer offered by the cloud layer to execute the storage service operation; transmitting, by the SSUI, an application programming interface (API) request for the SSUI agent for executing the storage service operation; and executing, by the deployed storage micro-service layer, the storage service operation.

Abstract: Methods and systems for a storage environment are provided. One method includes copying a data unit from a first temporary storage location corresponding to each zoned solid-state drive (ZNS SSD) of a first ZNS SSD set of a storage system to a first XOR module, while determining a first partial horizontal parity using the data unit stored in the first temporary storage location; and determining a vertical parity for each ZNS SSD of the first ZNS SSD set using the data unit provided to the first XOR module in a current cycle and vertical parity determined from a previous cycle.

Abstract: Multi-site distributed storage systems and computer-implemented methods are described for improving a resumption time of input/output (I/O) operations during a common snapshot procedure for storage objects. A computer-implemented method includes initiating a snapshot multi create operation to selectively form a batch of first and second synchronous replicated datasets that belong to a first group of storage disks at the primary storage site and corresponding second group of storage disks at the secondary storage site, performing a batch snapshot create operation on the primary storage site by executing snapshots of storage objects on the primary storage site of the batch of first and second synchronous replicated datasets in parallel multiple threads to effectively utilize processing resources on the primary storage site, and initiating an independent workflow and state machine for each storage object of the batch of first and second synchronous replicated datasets.

Abstract: Systems and methods for reducing read application in a virtual storage system are provided. According to one embodiment, read amplification is reduced when AZCS compression is being utilized by avoiding restarting of a read process via a slow path via a RAID layer of the virtual storage system when a data block associated with a read request and obtained via a first fast path read has been found not to be compressed. Instead, a second fast path read may be performed to obtain the corresponding checksum. Alternatively, or additionally, heuristics may be used to predict the odds of the data block being compressed. For example, when information encoded within a PVBN of the data block that identifies the PVBN as being within a compressed AZCS zone has shown to be sufficiently/insufficiently predictive of the data block being compressed, then a flag may be set to enable/disable fast path reads.

Abstract: Multi-site distributed storage systems and computer-implemented methods are described for providing common snapshot retention and automatic fanout reconfiguration for an asynchronous leg after a failure event that causes a failover from a primary storage site to a secondary storage site. A computer-implemented method comprises providing an asynchronous replication relationship with an asynchronous update schedule from one or more storage objects of the first storage node to one or more replicated storage objects of the third storage node, creating a snapshot copy of the one or more storage objects of the first storage node, transferring the snapshot copy to the third storage node based on an asynchronous mirror policy, and intercepting the snapshot create operation on the primary storage site and transferring the snapshot copy to the second storage node to provide a common snapshot between the second storage node and the third storage node.

Abstract: A data management system can include a set of storage media configured to implement a storage space and a set of controllers. The set of controllers can be configured to write to the storage space and to implement a set of nodes. The set of controllers can include a first controller that implements a first node and includes a first persistent memory, a second controller that implements a second node and includes a second persistent memory and a third controller that implements a third node and includes a third persistent memory. The third node can be configured to write third node journal data to the first persistent memory. The first node can be configured to generate first node journal data based on a first request received from a backend, write the first node journal data to the first persistent memory, and replicate the journal data to the second persistent memory.

Abstract: A distributed storage management system comprising nodes that form a cluster, a distributed block layer that spans the nodes in the cluster, and file system instances deployed on the nodes. Each file system instance comprises a data management subsystem and a storage management subsystem disaggregated from the data management subsystem. The storage management subsystem comprises a node block store that forms a portion of the distributed block layer and a storage manager that manages a key-value store and virtualized storage supporting the node block store. A file system volume hosted by the data management subsystem maps to a logical block device hosted by the virtualized storage in the storage management subsystem. The key-value store includes, for a data block of the logical block device, a key that comprises a block identifier for the logical block device and a value that comprises the data block.