Abstract: A backup system and method for a large data store employs a new and enhanced summary change log data structure that records all file system changes comprises inode entry records augmented with additional fields to indicate inodes representing related child and parent file system objects. A single preprocessing pass through the summary change log augments all of the appropriate inode fields with the appropriate inode numbers of family members to create a graph structure that can be traversed in hierarchical order starting at a root inode to quickly identify all created and modified file system objects for inclusion in an incremental backup.

Abstract: Various computer-implemented methods for selective defragmentation are disclosed. In one example, an exemplary method for performing such a task may include: 1) identifying a file system, 2) tracking changes to files in the file system, 3) creating a list of changed files, and then 4) selectively defragmenting files in the list of changed files. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Various systems and methods use advisory metadata to track the status of file system objects. For example, one method can involve accessing a first item of advisory metadata, which is associated with a namespace-visible parent directory and identifies a first transaction and a first time, and accessing a second item of advisory metadata, which is also associated with the namespace-visible parent directory and identifies a second transaction and a second time. The namespace visible parent directory includes multiple namespace-hidden directories. The method can then select a time value of the advisory metadata for the namespace-visible parent directory, based upon the first item and the second item. The method can be performed by a computing device implementing a file system.

Abstract: Systems and methods for inode use are presented. In one embodiment; an inode reuse method includes: receiving an indication of an operation that involves access to file related information; assigning an inode to the access; identifying one of a plurality of inode reuse scenarios for the inode; and making the inode available for reuse in accordance with the one of the plurality of inode reuse scenarios. In one embodiment, the one of the plurality of inode reuse scenarios is a relatively expedited reuse scenario. In one exemplary implementation, the relatively expedited inode reuse scenario is utilized if the inode is not required for further processing associated with the operation. The inode can be reused for another immediately subsequent operation.

Abstract: A system and method for repairing potentially corrupt file systems are provided. A request to repair an original file system, including a current set of metadata, is received. A metasave image is accessed, in response to the request. The metasave image includes a copy of the metadata included in the original file system at a point-in-time and information identifying a location of each block of metadata in the metasave image. Using the metasave image information, the current set of metadata in the original file system is replaced with the metadata in the metasave image.

Abstract: A computer-implemented method for migrating files to tiered storage systems may include 1) identifying at least one file to migrate from a first file system to a second file system, where the second file system may include a plurality of storage tiers, 2) identifying file metadata associated with the file as stored on the first file system, 3) before writing the file to the second storage system, selecting a storage tier from the plurality of storage tiers based at least in part on the file metadata, and 4) writing the file directly to the selected storage tier within the second file system to avoid moving the file to the selected storage tier after writing the file to the second file system. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A system and method for migrating file systems is provided. A file system migration operation copies a plurality of files from a first file system to a second file system. A request to access one of the files being migrated during the file system migration operation results in the creation of a pseudo vnode for the file being requested in the second file system. The pseudo vnode is created prior to creating a vnode for the file in the second file system. In addition, the pseudo vnode is associated with a vnode of the file at the first file system. Information identifying the pseudo vnode is then returned to the application.

Abstract: A method of partitioning directory. Accesses, e.g., shared/exclusive, and/or waiting requests, e.g., shared/exclusive, to access one or more files with a directory are monitored, e.g., incrementing/decrementing respective counters. The waiting requests are queued to be granted at a later time. The directory is determined to be primed for partitioning if a number of waiting requests to access the directory is greater than a threshold value of a plurality of heuristics and optionally further based on satisfying the condition for at least a programmable time threshold period. A trigger signal is automatically generated if the directory is primed for partitioning. The trigger signal causes a file system to partition the directory. It is appreciated that the plurality of heuristics is user programmable.

Abstract: A system and method for efficiently performing simultaneous access of files in a clustered file system. A cluster of nodes is coupled to a data storage medium. A first node executes a first thread with write operations for a given file. A second node executes a second thread with read operations for the given file. The first thread is appending the given file. In response, the first node requests a write lock range between an end-of-file for the given file to a highest possible endpoint. If a prediction determines the second thread may request read locks after the range it is currently reading, then the method enables the locks to be efficiently transferred from the first node to the second node before they are required on the second node, allowing the reading second node to proceed with fewer delays.

Abstract: A computer-implemented method for copying database files may include (1) identifying a request to copy a database file to a target location, and in response to identifying the request to copy the database file to the target location, (2) identifying a preallocated but unused portion of the database file, (3) transmitting a used portion of the database file to the target location, and (4) transmitting a metadata representation of the unused portion of the database file to the target location, the metadata representation being smaller than the unused portion. Various other methods and systems are also disclosed.

Abstract: A computer-implemented method for host communication in a clustered file system may include 1) identifying a cluster file system site comprising at least one host that performs data operations on a storage unit, 2) identifying a storage management application that manages the storage unit, 3) transmitting, by an initial host, an initial data operation directed to a location of the storage unit, where the data operation comprises a notification signal that commands the notification of the initial host upon a subsequent host attempting to access the location, 4) transmitting, by the subsequent host, a subsequent data operation directed to the location of the storage unit, and 5) notifying the initial host, by the storage management application, that the subsequent host attempted to access the location of the storage unit. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A computer-implemented method may include identifying a file that includes one or more data blocks stored within a storage device. The computer-implemented method may also include determining that the file has been accessed more frequently than one or more other files stored within the storage device and then tracking the number of times that at least one data block associated with the file is accessed in the storage device. The computer-implemented method may further include determining that the number of times that the data block associated with the file has been accessed in the storage device is above a predetermined threshold. In addition, the computer-implemented method may include caching the data block associated with the file in a caching mechanism to decrease the amount of time required to access the data block. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A method of updating a link count in a partitioned directory shared by a plurality of computers within a cluster. The partitioned directory is traversed once by a first computer to obtain a link count. The link count is stored in the first computer and broadcast to a cluster of computers. A total number of allocated links is less than a maximum limit on link count minus the link count. A respective number of links is allocated to each computer within the cluster. Delta values of each computer are updated in real-time as subdirectories are created/erased. A delta value associated with each of the plurality of computers is received. A delta value represents a net number of links created or removed by each computer. The link count is updated based on the previous link count and further based on each of the delta values and further based a delta value associated with the first computer. The updated link count is stored by the first computer and broadcast to the plurality of computers.

Abstract: A method for file lock recovery in a distributed computer system. The method includes executing a distributed computer system having a plurality of nodes comprising a cluster, and initiating a network file system server migration from one node of the cluster to a different node of the cluster. A migration count is incremented, wherein the migration count is stored at each of the nodes comprising the cluster. File lock services are paused at each of the nodes comprising the cluster. The network file system server migration is completed at the different node of the cluster. The migration count is then decremented in response to the completion. File lock services are then resumed at each of the nodes comprising cluster.

Abstract: Embodiments of the present invention are directed to a method and system for managing resources. The method includes receiving a request, within an electronic system, which corresponds to an object of a file system and accessing a local data structure. The data structure comprises information corresponding to a plurality of inode numbers. The method further includes performing the request and updating the local data structure based on the request. The updating of the local data structure is independent of a plurality of data structures of a cluster of servers.

Abstract: A computer-implemented method for performing file system checks may include identifying a file system to check. The computer-implemented method may also include identifying a base log including file system metadata within the file system at a previous time. The computer-implemented method may further include identifying a set of changed inodes within the file system since the previous time. The computer-implemented method may additionally include creating an updated log based on the base log and the set of changed inodes. The computer-implemented method may also include checking the updated log for consistency. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A method for fast incremental replication of a file system. The method includes, at a data storage level, tracking file system status for a plurality of files by using a data structure inside respective inodes for the plurality of files, and at a file system level, tracking file system status for a plurality of files by using a file system mask structure. For all files present in a backup, a catalogue of the file path name and inode number is maintained on a source file system. During incremental backup, a source file system data structure enumerating a plurality of modified inodes is consulted. For inodes that have not been marked as removed or created, the file system mask structure that tracks modified portions is consulted to obtain a file incremental change and inodes. The method further includes performing incremental backup using the inodes that have been modified and appended.

Abstract: A method for fast incremental replication of a file system. The method includes, at a data storage level, tracking file system status for a plurality of files by using a data structure inside respective inodes for the plurality of files, and at a file system level, tracking file system status for a plurality of files by using a file system mask structure. For all files present in a backup, a catalogue of the file path name and inode number is maintained on a source file system. During incremental backup, a source file system data structure enumerating a plurality of modified inodes is consulted. For inodes that have not been marked as removed or created, the file system mask structure that tracks modified portions is consulted to obtain a file incremental change and inodes. The method further includes performing incremental backup using the inodes that have been modified and appended.

Abstract: A method for file lock recovery in a distributed computer system. The method includes executing a distributed computer system having a plurality of nodes comprising a cluster, and initiating a network file system server fail over from one node of the cluster to an adoptive node of the cluster. File lock services are then stopped at the adoptive node. File lock services are subsequently restarted at the adoptive node, wherein the restart causes the adoptive node to commence a grace period for other network file system clients to connect to the adoptive node and reclaim file locks. After restarting file lock services, a cluster file system is updated on the adoptive node with process identifiers, and file lock services are resumed at the adoptive node. The cluster file system can be simultaneously exposing the file lock services to other clients as well, like CIFS, etc.

Abstract: A computer-implemented method for migrating virtual machines may include: 1) identifying a request to migrate a virtual machine from a primary site to a secondary site, the primary site including a primary storage device used by the virtual machine and configured for asynchronous replication to a secondary storage device at the secondary site, 2) identifying a difference map that reflects differences between data on the primary storage device and the secondary storage device, 3) initiating the virtual machine at the secondary site, 4) intercepting an input/output attempt from the virtual machine at the secondary site to the secondary storage device, 5) determining, based on the difference map, that a region of the input/output attempt has not been synchronized from the primary storage device, and 6) retrieving the region from the primary storage device before allowing the input/output attempt to proceed. Various other methods, systems, and computer-readable media are disclosed.