Abstract: A method for managing a plurality of computing machines includes accessing a catalogue memory structure storing a plurality of component signatures, discovering one or more of the software components being instantiated in a software image of each computing machine according to corresponding ones of the component signatures, detecting one or more connections each being established between at least two instantiated software components of different computing machines, receiving a restore command for restoring a target recovery point selected among a plurality of recovery points, and restoring the target recovery point in response to the restore command by restoring the snapshots associated with the target recovery point on corresponding computing machines.

Abstract: A method for performing backups. The method includes sending, via a data manager, a quiesce request to a client storage array of a client device to put the client storage array in a quiesced state. The method also includes obtaining, via a primary backup storage system, a snapshot-based backup of a storage volume of the client storage array while the client storage array is in the quiesced state. Further, the method includes sending an unquiesce request to the client storage array in response to determining that the obtaining of the snapshot-based backup is complete. Moreover, the method includes sending the snapshot-based backup from the primary backup storage system to a secondary backup storage system. In addition, the method includes generating, using the secondary backup storage system, a protection copy based on the snapshot-based backup.

Abstract: A data management system (DMS) may support the backing up of scalable computing objects, such as hierarchical computing objects. For example, the DMS may backup a computing system and facilitate that capture and storage of snapshots of the computing system. The DMS may determine to capture a first snapshot of a computing object within the computing system. The DMS may identify components corresponding to respective portions of the computing object and cause the computing system to generate respective snapshots of the components. That is, the DMS may cause the computing system to generate multiple individual snapshots that together may constitute snapshot of the computing object. The DMS may generate the first snapshot that includes metadata for the snapshots of the components. The metadata may include a first identifier for the first snapshot and respective identifiers for the respective snapshots of the components.

Abstract: In some embodiments, a process for restoring a version of a virtual machine using a data storage system comprises identifying a particular version of a virtual machine to be restored, determining a base image from which the particular version may be derived, determining a set of incremental files for generating the particular version, generating a file associated with the particular version using the base image and the set of incremental files, and outputting at least a portion of the file.

Abstract: Disclosed is a method, a device, a system and/or a manufacture of immutable protection of software and/or computing hardware testing data. In one embodiment, a method for immutably securing testing data includes initiating an operation filesystem to operate from and copying a software application and/or computer hardware design to define a workspace data. A test script is selected and executed. The method provides input data including the testing data to a hash function. The testing data may include the result data. A hash value from the hash function is generated that is uniquely determined based on the input data. A blockchain transaction is generated that includes the hash value. The hash value is submitted to a node of a blockchain network that includes a distributed network of two or more nodes maintaining a ledger database reconciled through a consensus mechanism to immutably protect the testing data.

Abstract: Techniques described herein relate to methods and systems for managing backup operations. The method may include receiving a request to perform a first backup operation for a first virtual machine (VM); making a first determination, using a vProxy preference map, that a first vProxy is assigned to the first VM based on a backup capability associated with the first vProxy; making a second determination that the first vProxy is not available to perform the first backup operation; making a third determination, based on the second determination and using a vProxy information database, that a second vProxy is available to perform the first backup operation based on the second vProxy being associated with the same backup capability as the first vProxy; and performing the first backup operation using the second vProxy.

Abstract: A method for system recovery may be provided. The method may be implemented on a management server of a recovery system. The recovery system may further include at least one working server configured to perform one or more tasks and a global storage device configured to store a backup system or a system image file of each of the at least one working server. The management server may be connected to the at least one working server and the global storage device. The method may include, for each of the at least one working server, determining whether the working server is in an abnormal state. For each of the at least one working server, in response to determining that the working server is in an abnormal state, the method may further include assigning at least part of the global storage device to the working server and causing the working server to perform a first system recovery operation using the at least part of the global storage device.

Abstract: Systems and methods for generating container images are described. An image definition file specifying a set of packages is received at a virtual image server. The image definition file may specify a set of packages and the virtual image server comprises a mapping of the image definition file to an image file corresponding to each of the set of packages. For each of the set of packages, the corresponding image file may be retrieved from an image registry and used to generate, by the virtual image server, an image manifest that references the corresponding image file for each of the set of packages as a separate image layer. The image manifest is provided to a container host, which may build a container image based on the image manifest.

Abstract: A storage device includes a nonvolatile memory device and a memory controller. After writing first data at the first page, the memory controller writes second data at a second page. The memory controller generates first check data corresponding to the first data and second check data corresponding to the second data. The memory controller recovers the first and second physical addresses and the first and second logical addresses based on the second check data.

Abstract: A temporal graph model is provided that can handle all six graph updates cases: vertex created; vertex deleted; vertex property updated; edge created; edge deleted; and edge property updated. A memory efficient temporal graph storage solution stores the delta change of graph evolution at each update (e.g., instead of storing the whole graph snapshot at each time). Temporal graph querying techniques are described that can determine what the graph looks like at a given time in the past and can perform any graph query on that graph's snapshot at that time as if time traveling back to that time. Also provided are event sourcing techniques of recording and tracing temporal graph updates that make the evolution of the temporal graph explainable. Further provided is a seamless integration with a massive parallel graph processing and storage engine that fully leverages the multicore and multinode cluster computation and storage resources.

Abstract: Techniques are provided for a snapshot difference interface integrated into an object store data management container. The snapshot difference interface is capable of interpreting an object format and snapshot file system format of snapshots backed up to an object store within objects formatted according to the object format. The snapshot difference interface can identify differences between snapshots, such as files that changed between the snapshots, while the snapshots are still resident within the object store. Because the snapshot difference interface does not retrieve the snapshots from the object store, security is improved, resource and network consumption is reduced, there is less of an impact upon client I/O processing, and a catalog of the snapshots can be more efficiently built and recovered in the event of corruption.

Abstract: Methods and systems for generating a computer program for a governance, risk, compliance (GRC) system are described. In an example, a processor may detect a change among a plurality of objects in the GRC system. The processor may generate a snapshot of the plurality of objects in response to the detected change. The snapshot may include the detected change among the plurality of objects. The processor may train a machine learning model using the snapshot. The trained machine learning model may indicate relationships among a set of related objects, where the related objects may be among the plurality of objects. The processor may generate a computer program based on the trained machine learning model. The computer program may include a set of instructions for setting a field value of an object among the set of related objects.

Abstract: An online system performs upgrades of replicated databases. The online system stores data in a primary site and a secondary site. The primary site comprising databases used to process client requests and a secondary site comprising databases used as standby. The online system upgrades databases of both sites by upgrading the secondary site, switching client requests to the secondary site while using the databases of primary site as standby databases. The online system then upgrades the primary site. The process allows the online system to upgrade databases with small down time.

Abstract: Disclosed are a computer-implemented method, a system and a computer program product for record management in a time series database (TSDB). In the computer-implemented method for record management in a TSDB, time series records can be obtained from distributed sources by one or more processing units. The records can be divided into different record groups based on a timestamp of each of the records by one or more processing units. The records in each of the record groups can be normalized into normalized records by one or more processing units. The normalized records can be stored in the TSDB by one or more processing units.

Abstract: According to one aspect, data collected from various modules on an autonomous vehicle is synchronized, and snapshots of the synchronized data, referred to herein as ‘data snapshots’, are created. Contextual information associated with the data snapshots, e.g., indexes and/or metadata, is collected and/or derived, and the contextual information is stored along with the snapshots. The data snapshots may then be substantially searched, using the contextual information, to identify particular scenes for analysis. In addition to searching for particular scenes, series of scenes may be searched using the contextual information such that scenarios encountered by the autonomous vehicle over time may be identified.

Abstract: Techniques are provided for forwarding operations to bypass persistent memory. A modify operation, targeting an object, may be received at a persistent memory tier of a node. If a forwarding policy indicates that forwarding is not enabled for the modify operation and the target object, then the modify operation is executed through a persistent memory file system. If the forwarding policy indicates that forwarding is enabled for the modify operation and the target object, then the modify operation is forwarded to a file system tier as a forwarded operation for execution through a storage file system.

Abstract: Dynamic snapshot scheduling techniques are provided using storage system metrics. One method comprises obtaining a schedule for generating snapshots of a portion of a storage system; automatically adjusting snapshot generation parameters in the schedule based on: (i) a current storage pool usage metric, (ii) an input/output metric of at least one storage resource in the portion of the storage system, (iii) a measure of snapshots in a destroying state, and/or (iv) a measure of a number of created snapshots; and initiating a generation of a snapshot of the storage system portion in accordance with the adjusted schedule. A snapshot generation frequency may be increased in response to an increase of: the current storage pool usage metric, the number of snapshots in the destroying state, and/or the number of created snapshots. A snapshot generation frequency may be decreased in response to an increase of the I/O metric of the at least one storage resource.

Abstract: A container image is received, the container image including a container manifest referencing a first base operating system layer and an application layer. An execution compatibility is determined between the application layer and a second base operating system layer, different from the first operating system layer. An updated container image, an updated container manifest, or the updated container image and the updated container manifest is generated in which the first base operating system layer is replaced with the second base operating system layer.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: Embodiments disclosed are directed to a computing system that provides for on-demand scanning for changes in cloud-based object storage systems. The changes can include modifications, additions, or deletions of objects stored in a cloud-based object storage system. The computing system scans, at a first runtime, objects stored in a cloud-based object storage system. The computing system retrieves lookup information associated with a second runtime of a previous scan. Based on the lookup information, the computing system identifies a subset of the objects that were changed or added after the second runtime and before the first runtime. Subsequently, the computing system generates an electronic notification including a consolidated list of the subset of the objects.

Abstract: Generation of new snapshots of a storage object are prompted based on changes to the storage object. An absolute storage change condition is satisfied when the aggregate amount of changed data since the most recent snapshot satisfies a threshold that represents a predefined amount of storage capacity, e.g., N GB. A percentage of used storage change condition is satisfied when the aggregate size of the changes expressed as a percentage of the provisioned storage capacity of snapped storage object satisfies a predetermined threshold, e.g., X % of Y TB. A percentage of provisioned storage change condition is satisfied when the aggregate amount of the changes in terms of percentage of utilized storage capacity of storage object satisfies a predetermined threshold expressed as a percentage of the utilized capacity of the storage object, e.g., P % of Z TB.

Abstract: One example method includes refreshing secondary or repurposed copies. A copy data management system issues a single request that allows multiple secondary copies to be refreshed using an API server. The server allows a storage controller to read data from a single source and then write the data read from the source to multiple targets simultaneously, thereby refreshing multiple copies in parallel.

Abstract: A method of consolidating snapshots includes receiving a request to consolidate a first snapshot with a second snapshot into a third snapshot, the first and second snapshots stored in separate backup files, each backup file organized as a directory where data parts of the first and second snapshots can be hard linked to locations outside of the backup file, comparing the data parts of the first and second snapshots to determine if any second snapshot data part fully overlaps with any first snapshot data part, responsive to determining that a second snapshot data part fully overlaps with a first snapshot data part, hard linking the determined second snapshot data part into the third snapshot, and storing the third snapshot in the backup file.

Abstract: In accordance with some aspects of the present disclosure, a non-transitory computer readable medium is disclosed. In some embodiments, the non-transitory computer readable medium includes instructions that, when executed by a processor, cause the processor to receive, from a workload hosted on a host of a cluster, first I/O traffic programmed according to a first I/O traffic protocol supported by a cluster-wide storage fabric exposed to the workload as being hosted on the same host. In some embodiments, the workload is recovered by a hypervisor hosted on the same host. In some embodiments, the non-transitory computer readable medium includes the instructions that, when executed by the processor, cause the processor to adapt the first I/O traffic to generate second I/O traffic programmed according to a second I/O traffic protocol supported by a repository external to the storage fabric and forward the second I/O traffic to the repository.

Abstract: Contents of a virtual machine of a primary storage are received at a first point in time. A full snapshot of the virtual machine is stored on a secondary storage using the received contents of the virtual machine. One or more cloud storage volumes of the virtual machine are caused to be created. Contents of the full snapshot for storage on the one or more cloud storage volumes are provided from the secondary storage. Changed data of the virtual machine of the primary storage is received at a second point in time. An incremental snapshot of the virtual machine is stored using the received changed data. The one or more cloud storage volumes are updated using the incremental snapshot of the virtual machine. An image of at least one of the cloud storage volumes is caused to be created for standby deployment of another instance of the virtual machine.

Abstract: A computer-implemented method according to one embodiment includes receiving, on a first cluster site, a first I/O request to migrate a plurality of filesets from a second cluster site to the first cluster site. The first cluster site includes a plurality of gateway nodes. The method further includes identifying at least two of the gateway nodes having resources available to perform operations of the migration, and hashing information of a plurality of filesets against the identified gateway nodes. The information includes inode numbers of entities that are mounted during fulfillment of the first I/O request. Operations of the first I/O request are distributed to the identified gateway nodes based on the hashing, and the identified gateway nodes are instructed to fulfill the operations.

Abstract: A cloud manager restores a backup of a distributed environment by comparing a stack of a current distributed environment to a backup stack of the backup of the distributed environment. The cloud manager identifies a difference between a first number of nodes in the first set of nodes specified in the first stack and a second number of nodes in the second set of nodes specified in the second stack. The cloud manager restores the backup of the distributed environment by modifying a number of nodes in the current distributed environment based on the identified difference between the first number of nodes and the number of nodes in the second stack.

Abstract: Examples described herein generally relate to maintaining a snapshot of a multitier distributed file system. A snapshot directory volume of a directory volume can be created at least in part by copying, to the snapshot directory volume, metadata of the directory volume. Based on a file access request, it can be determined whether the file has not been snapshotted. If not, a snapshot file record of the file, and associated with the snapshot directory volume.

Abstract: A virtual storage volume that includes storage space on potentially many different physical disks may be implemented as one or more chunks. This allows for dynamic volume storage adjustment. Dynamic volume storage adjustment allows for increasing volume size without moving data unnecessarily. In addition, dynamic volume storage adjustment also allows freeing up storage space in a node with minimum movement of data.

Abstract: Implementations described herein relate to methods, systems, and computer-readable media to selectively synchronize database objects. In some implementations, a method includes receiving a request to synchronize records that include a corresponding root record and one or more connected records from a first database instance to a second database instance, wherein the request includes a specified graph configuration and an object boundary identifier; collecting the one or more records by traversing a corresponding graph configuration associated with each of the one or more root records; generating a snapshot that includes the one or more records into one or more object record sets based on a corresponding object boundary identifier associated with each of the one or more root records and the one or more connected records; publishing the snapshot to the second database instance; and updating the second database instance.

Abstract: Provided are techniques for role reversal of primary and secondary sites with minimal replication delay. During a backup phase, an Input/Output (I/O) operation selected from a group comprising a create operation and a rename operation is received with one or more local attributes of a local file. The I/O operation is processed to generate one or more remote attributes. The one or more local attributes are stored with the one or more remote attributes. During a reversed phase, a write operation for the remote file is processed, and, the write operation is sent to the primary replication system with the one or more local attributes, where the primary replication system uses a local file handle in the one or more local attributes to access the local file for processing the write operation to sync the local file with the remote file.

Abstract: A system receives a request to import data file(s) from a source data store into a target database. The system reserves a first portion of computing resources that host the target database to import the data file(s). The reservation of the first portion of computing resources permits the import throughput rate of the data file(s) through the first portion of computing resources while maintaining a second portion of the computing resources to support client access to the target database at an access throughput rate. The system initiates import of the data file(s) from the source data store to the target database through one or more storage nodes at the import throughput rate according to the first portion of computing resources. The target database is able to receive access requests from one or more clients during the import of the data file(s) to the target database.

Abstract: Presented herein are systems and methods for backing up and restoring related backup data stored in different cloud storage tiers in a cloud storage environment. Backup data generated during a backup job is stored to different cloud storage tiers based on the type of data and a storage policy comprising designations of a first cloud storage tier and a second cloud storage tier. Backup metadata and index data is stored in the designated first cloud storage tier and backup payload data is stored in the designated second cloud storage tier. Designations of the storage tiers are based on the attributes associated with each tier of cloud storage. During restore, the backup payload data is recalled from the second storage tier into the first storage tier where it is used to restore to the destination client or customer.

Abstract: Distributed application orchestration management is provided. A first passive member of a set of passive members sends a notification message to other members indicating that the first passive member is initiating start of a distributed application in response to the first passive member validating that a self-restart by a leader member failed. The first passive member compares timestamps associated with an attempt to start the distributed application by other passive members in the set of passive members. The first passive member stops a particular attempt to start the distributed application in response to the first passive member determining that a timestamp associated with the particular attempt to start the distributed application by the first passive member is newer than another timestamp of another passive member. The first passive member designates the other passive member having an older timestamp as a new leader member to continue starting the distributed application.

Abstract: A client computing device receives a behavior model corresponding to a user group associated with a user. The behavior model has been trained with monitored user interactions of one or more files associated with the user group. The client computing device further mounts a snapshot of a file and determines, based on the behavior model, which files of the mounted snapshot to transfer to a locally accessible cache. During use of the client computing device, the client computing device may determine whether the mounted snapshot is accessible. If the mounted snapshot is not accessible, the client computing device may selectively delete, based on the behavior model, one or more of the files stored in the locally accessible cache. If the mounted snapshot is accessible, the client computing device may update the one or more files of the locally accessible cache with monitored user interactions with the mounted snapshot.

Abstract: A system and method for exposing volumes with underlying read-write mediums to user operations. When a medium is in the process of being migrated to a storage array, a volume which relies on the medium can be exposed to user operations (e.g., snapshots, read and write operations) once the portions of the medium which underlie the volume have been migrated. The volume can be exposed to user operations while one or more other portions of the medium are unfilled and while the medium is in an intermediate read-write state.

Abstract: A method, computer program product, and computer system for triggering, by a computing device, a synchronization activity on a storage object. A request may be executed for a copy operation resulting from the synchronization activity to copy the storage object from a source to a destination when a current number of concurrent copy operations is below a maximum number of copy operations. The request for the copy operation resulting from the synchronization activity to copy the storage object from the source to the destination may be placed in a queue when the current number of concurrent copy operations is at least at the maximum number of copy operations. One or more attributes of the request in the queue may be identified. The request for the copy operation may be executed according to a replication policy based upon, at least in part, the one or more attributes of the request in the queue.

Abstract: An illustrative embodiment disclosed herein is an apparatus including a processor having programmed instructions to detect an operation associated with a first object instance in a first cluster, determine that the first object instance is configured for replication to a second cluster, create an object information (info) metadata entry for the first object instance, create an object replication metadata entry for the first object instance, and simultaneously write the object info metadata entry and the object replication metadata entry to a first data structure and a second data structure, respectively.

Abstract: A lightweight always-on monitoring, collecting, diagnosing, and correcting utility operates in an enhanced storage manager that manages a data storage managements system. The always-on utility provides a comprehensive and pro-active approach, which is intended to reduce, if not altogether eliminate, the need for after-the-fact diagnostics. The always-on utility also enforces so-called best practices and other heuristics, which include pro-actively activating certain database settings that are not enabled by default; manipulating certain aspects of the database to improve performance; and reporting aspects that are outside best-practice parameters to the trouble report system so that system administrators and/or developers may intervene before a catastrophic failure occurs. In some cases, the best-practice parameters represent heuristics designed by the present inventors to improve the performance and general health of the management database.

Abstract: An agent executing on a processor at a data production site detects filesystem events and creates a list of bookmarks describing the filesystem events in user-understandable strings. Each bookmark is associated with metadata about the filesystem event that includes a timestamp and a point-in-time (PiT) snapshot of the filesystem at the time of occurrence of the event. The bookmarks and associated metadata are transferred to a recovery site and the bookmark set is recorded in storage. The storage can be searched for bookmarks in the set that reference the filename of a file of interest, and the file recovered and replicated using an associated PiT.

Abstract: Disclosed herein are methods, systems, and apparatus, including computer programs encoded on computer storage media, for managing transaction requests in ledger systems.

Abstract: Disclosed herein are methods, systems, and apparatus, including computer programs encoded on computer storage media, for avoiding double-spending problem in read-write set-model-based blockchain technology. One of the methods includes receiving instructions to execute two or more blockchain transactions on a piece of data, where all blockchain transactions of the two or more blockchain transactions modify a value of the piece of data, and for each blockchain transaction from the two or more blockchain transactions, pre-executing a smart contract associated with the blockchain transaction to generate a special instruction indicating the blockchain transaction, where the special instruction is used to validate that a current value of the piece of data supports the blockchain transaction when executing the smart contract to write the blockchain transaction to a blockchain.

Abstract: In an embodiment, a computing system may restore data for a protected system using an emulated block device that presents a synthesized filesystem. Initially, a request to restore data for the protected system may be received. The request may include or otherwise indicate a particular file system format, a particular virtual disk format, and a recovery point representing a particular point in time for presentation of the restored data. Responsive and subsequent to receiving the request to restore the data, the computing system generates a first configuration of the emulated block device for presenting the synthesized filesystem, the first configuration being based on: a) an amount of data in the recovery point, b) the particular file system format, and c) the particular virtual disk format. The computing system also generates a configuration for the synthesized filesystem, and provides access to the emulated block device based on the determined configurations.

Abstract: In a method used for maximizing performance of a storage system, saturation points of the storage system for I/O requests of different types and sizes are identified. Normalized tokens are determined based on the saturation points. Unique numbers of normalized tokens are associated with the I/O requests of different types and sizes. A number of normalized tokens for a queue depth limit of a storage device is determined. From the queue depth limit, normalized tokens are allocated to each node in the storage device. I/O requests are processed according to the allocations of normalized tokens.

Abstract: A system and method determining a Chain Identification Number (CID) of a source snapshot to be replicated from a source site to a target site of a virtual computing system, determining a predetermined number of potential reference snapshots based on the CID of the source snapshot, computing a closeness value between the source snapshot and each of the potential reference snapshots, and creating a list of the potential reference snapshots based on the closeness value of each of the potential reference snapshots. One snapshot from the list is selected as a reference snapshot. The source snapshot is replicated to the target site based on the reference snapshot.

Abstract: Described herein is a system that restores a database by processing a portion of the database. The system restores the database to a previous state at a particular time by reverting data entries that have changed since the time to their initial values before the change. Data entries that have changed after the restore time are identified. For the data entries that have changed after the restore time, their initial values before the change are determined from various sources. The system determines a database version that is created most recently before the restore time. The system additionally identifies changes to the database between the restore time and when the database version is created. The initial values can be determined from either the database version or the changes made to the database between the restore time and when the database version is created.

Abstract: Methods and systems for protecting virtual machines is provided. One method includes identifying a plurality of volumes of a primary storage system for storing a plurality of data containers of a virtual machine (“VM”); generating a consistency group (“CG”) snapshot of the plurality of volumes; associating a single instance storage clone of each data container of the VM from the CG snapshot to an active file system of the primary storage system; and generating metadata for representing a VM backup as a set having each of the single instance storage clone of each data container.

Abstract: Contents of a virtual machine of a primary storage are received at a first point in time. A full snapshot of the virtual machine is stored on a secondary storage using the received contents of the virtual machine. One or more cloud storage volumes of the virtual machine are caused to be created. Contents of the full snapshot for storage on the one or more cloud storage volumes are provided from the secondary storage. Changed data of the virtual machine of the primary storage is received at a second point in time. An incremental snapshot of the virtual machine is stored using the received changed data. The one or more cloud storage volumes are updated using the incremental snapshot of the virtual machine. An image of at least one of the cloud storage volumes is caused to be created for standby deployment of another instance of the virtual machine.

Abstract: System and method for managing snapshots of storage objects in a storage system use a consolidation operation to reduce read amplification for stored snapshots of a storage object that are stored in log segments in the storage system according to a log-structured file system as storage service objects. The consolidation operation involves identifying target log segments among the log segments that include live blocks that are associated with the latest snapshot of the storage object and determining the number of the live blocks included in each of the target log segments. Based on the number of the live blocks in each of the target log segments, candidate consolidation log segments are determined from the target log segments. The live blocks in the candidate consolidation log segments are then consolidated to new log segments, which are uploaded to the storage system as new storage service objects.

Abstract: Systems and methods for adding backups to de-duplicated storage and for removing backups from de-duplicated storage are disclosed. Backups can be added to the de-duplicated storage with minimal reference count updates. Backup data sets can be removed without garbage collection processes being performed on the de-duplicated storage.