Abstract: A data storage system includes multiple head nodes and multiple data storage sleds mounted in a rack. For a particular volume or volume partition one of the head nodes is designated as a primary head node for the volume or volume partition. The primary head node is configured to store data for the volume in a data storage of the primary head node and cause the data to be replicated to a secondary head node. The primary head node is also configured to cause the data for the volume to be stored in a plurality of respective mass storage devices each in different ones of the plurality of data storage sleds of the data storage system.

Abstract: Embodiments are directed to processing read-only operations without storing the operations persistently, determining a processing order for read-only requests, and to directing read-only requests to nodes best able to handle the requests. In an embodiment, a quorum leader receives various read-only requests. The quorum leader controls process ordering for replicas in the leader's quorum. The quorum leader determines an appropriate order for processing the read-only operations, without clock-based synchronization. The quorum leader sends the read-only request to an appropriate replica in the quorum and executes the read-only request at the appropriate replica without storing the operation persistently.

Abstract: An opportunistic throttling approach can be used for customers of shared resources in a multi-tenant environment. Each customer can have a respective token bucket with a guaranteed fill rate. When a request is received for an amount of work to be performed by a resource, the corresponding number of tokens are obtained from, or charged against, a global token bucket. If the global bucket has enough tokens, and if the customer has not exceeded a maximum work rate or other such metric, the customer can charge less than the full number of tokens against the customer's token bucket, in order to reduce the number of tokens that need to be taken from the customer bucket. Such an approach can enable the customer to do more work and enable the customer's bucket to fill more quickly as fewer tokens are charged against the customer bucket for the same amount of work.

Abstract: An opportunistic throttling approach can be used for customers of shared resources in a multi-tenant environment. Each customer can have a respective token bucket with a guaranteed fill rate. When a request is received for an amount of work to be performed by a resource, the corresponding number of tokens are obtained from, or charged against, a global token bucket. If the global bucket has enough tokens, and if the customer has not exceeded a maximum work rate or other such metric, the customer can charge less than the full number of tokens against the customer's token bucket, in order to reduce the number of tokens that need to be taken from the customer bucket. Such an approach can enable the customer to do more work and enable the customer's bucket to fill more quickly as fewer tokens are charged against the customer bucket for the same amount of work.

Abstract: A block-based storage system may implement page cache write logging. Write requests for a data volume maintained at a storage node may be received at a storage node. A page cache for may be updated in accordance with the request. A log record describing the page cache update may be stored in a page cache write log maintained in a persistent storage device. Once the write request is performed in the page cache and recorded in a log record in the page cache write log, the write request may be acknowledged. Upon recovery from a system failure where data in the page cache is lost, log records in the page cache write log may be replayed to restore to the page cache a state of the page cache prior to the system failure.

Abstract: Embodiments are directed to processing read-only operations without storing the operations persistently, determining a processing order for read-only requests, and to directing read-only requests to nodes best able to handle the requests. In an embodiment, a quorum leader receives various read-only requests. The quorum leader controls process ordering for replicas in the leader's quorum. The quorum leader determines an appropriate order for processing the read-only operations, without clock-based synchronization. The quorum leader sends the read-only request to an appropriate replica in the quorum and executes the read-only request at the appropriate replica without storing the operation persistently.

Abstract: Embodiments are directed to processing read-only operations without storing the operations persistently, determining a processing order for read-only requests, and to directing read-only requests to nodes best able to handle the requests. In an embodiment, a quorum leader receives various read-only requests. The quorum leader controls process ordering for replicas in the leader's quorum. The quorum leader determines an appropriate order for processing the read-only operations, without clock-based synchronization. The quorum leader sends the read-only request to an appropriate replica in the quorum and executes the read-only request at the appropriate replica without storing the operation persistently.

Abstract: The present invention extends to methods, systems, and computer program products for implementing a cache using multiple page replacement algorithms. An exemplary cache can include two logical portions where the first portion implements the least recently used (LRU) algorithm and the second portion implements the least recently used two (LRU2) algorithm to perform page replacement within the respective portion. By implementing multiple algorithms, a more efficient cache can be implemented where the pages most likely to be accessed again are retained in the cache. Multiple page replacement algorithms can be used in any cache including an operating system cache for caching pages accessed via buffered I/O, as well as a cache for caching pages accessed via unbuffered I/O such as accesses to virtual disks made by virtual machines.

Abstract: A method for establishing and maintaining a differential snapshot of a set of files stored on a volume is disclosed. The invention achieves processing time and disk space optimizations by avoiding copy-on-write operations for logically insignificant moves of blocks, such as the block rearrangements characteristic of defragmentation utilities. A file system enhancement enabling the passing of a block copy command from the file system to lower-level drivers, is used to inform the snapshotter that a block move operation is not logically meaningful. When the logically insignificant move is of a block whose data forms part of the data captured in the snapshot virtual volume, and when the move is to a block location that is functioning as logical free space, the snapshotter can simply modify its block bitmap and update translation table entries without needing to perform a copy-on-write.

Abstract: Embodiments are directed to processing read-only operations without storing the operations persistently, determining a processing order for read-only requests, and to directing read-only requests to nodes best able to handle the requests. In an embodiment, a quorum leader receives various read-only requests. The quorum leader controls process ordering for replicas in the leader's quorum. The quorum leader determines an appropriate order for processing the read-only operations, without clock-based synchronization. The quorum leader sends the read-only request to an appropriate replica in the quorum and executes the read-only request at the appropriate replica without storing the operation persistently.

Abstract: Aspects of the subject matter described herein relate to transactions. In aspects, a consistency requester such as a shadow copy service requests to have a consistent view of a set of resources. In response, transactions that are in a problematic state are allowed to exit the problematic state while new and existing transactions that are not in the problematic state are allowed to perform any work except work that takes them into the problematic state. After no transactions are in the problematic state, a consistent view of the set of resources is available. This view may be used by the consistency requester as desired. Afterwards, the consistency requester may indicate that a consistent view is no longer needed. Transactions are then allowed to enter the problematic state. Transactions may also be allowed to enter the problematic state if a timeout elapses.

Abstract: A method for establishing and maintaining a differential snapshot of a set of files stored on a volume is disclosed. The invention achieves processing time and disk space optimizations by avoiding copy-on-write operations for logically insignificant moves of blocks, such as the block rearrangements characteristic of defragmentation utilities. A file system enhancement enabling the passing of a block copy command from the file system to lower-level drivers, is used to inform the snapshotter that a block move operation is not logically meaningful. When the logically insignificant move is of a block whose data forms part of the data captured in the snapshot virtual volume, and when the move is to a block location that is functioning as logical free space, the snapshotter can simply modify its block bitmap and update translation table entries without needing to perform a copy-on-write.

Abstract: A differential snapshot is established and maintained for a set of files stored on a volume. Copy-on-write operations are avoided for logically insignificant moves of blocks, such as the block rearrangements characteristic of defragmentation utilities. A file system passes a block copy command to lower-level drivers that are to inform the snapshotter that a block move operation is not logically meaningful. When the logically insignificant move is of a block whose data forms part of the data captured in the snapshot virtual volume, and when the move is to a block location that is functioning as logical free space, the snapshotter can simply modify its block bitmap and update translation table entries without needing to perform a copy-on-write.

Abstract: Volume configuration data management systems and methods are described. A logical volume provider maintains an epoch number on each of the extents comprising a volume. Upon each change in the configuration of the volume, the epoch number on each of the currently online extents is updated. When a volume is discovered and a request is made to bring the volume online, the logical volume provider can compare the epoch numbers on the extents to determine if the volume data is consistent across each of the extents, or if an extent contains stale data due to a transitory outage.

Abstract: Described is a revert preview mechanism, which gives the user a chance to temporarily remount a storage volume with the contents of a shadow copy, and then evaluate the computer system as if reverted, including allowing full I/O read and write operations to the storage volume. A temporary storage location is provided to cache data writes. Read requests are read back from the temporary storage location if the corresponding volume location has been written, or are read back from the shadow copy of the volume if not written since the revert preview operation began. In the case of a revert preview of the boot volume, an API is used to specify a correct shadow copy instance of the system registry hive, which needs to be loaded following reboot, but before the shadow copy driver is loaded. Use of the API loads the correct instance on the next reboot.

Abstract: Systems and methods for performing an on-demand determination of a correct set of volumes to be exported are described. A disk platter is defined to be the unit of volume migration. Volumes are added to the platter based on whether or not the volume coexists on a disk containing a volume to be exported.

Abstract: A mount manager and supporting data structures enable automatic identification and re-establishment of logical volumes on non-removable storage devices in a computer system across multiple reboots and reconfigurations. The mount manager generates a redirected name for a new logical volume when a unique volume identifier is presented to the mount manager by the operating system. The mount manager stores the unique volume identifier and the associated redirected name in a persistent mount manager data structure. The mount manager establishes a symbolic link between the persistent redirected name and a non-persistent device name used by the operating system. During the boot process, the mount manager uses the data structure entries identified by the unique volume identifiers of the arriving logical volumes to reconstruct the symbolic links so that references to the redirected name will resolve to the correct non-persistent device name.

Abstract: A storage management system in which a volume provider incorporates an application programming interface (API) for receiving storage access information that characterizes high-level input/output (I/O) attributes of the storage volume such as intended access patterns, a desired level of fault handling and optimization preferences. The volume provider and the underlying storage devices conform to the specified attributes as best as possible and resolve any conflicts with the communicated attributes. The volume providers monitor the storage volumes and may reconfigure the storage volumes in order to accomplish the desired attributes.

Abstract: Volume configuration data management systems and methods are described. A logical volume provider maintains an epoch number on each of the extents comprising a volume. Upon each change in the configuration of the volume, the epoch number on each of the currently online extents is updated. When a volume is discovered and a request is made to bring the volume online, the logical volume provider can compare the epoch numbers on the extents to determine if the volume data is consistent across each of the extents, or if an extent contains stale data due to a transitory outage.

Abstract: Systems and methods for performing an on-demand determination of a correct set of volumes to be exported are described. A disk platter is defined to be the unit of volume migration. Volumes are added to the platter based on whether or not the volume coexists on a disk containing a volume to be exported.