Microsoft Technology Licensing, LLC

Vishal Kathuria

Hanumantha Rao Kodavalla

Robin Dhananjay Dhamankar

Martin J Sleeman

Michael James Zwilling

Robert H Gerber

Yixue Zhu

Lev Novik

Rui Wang

Tomas Talius

Yunxin Wu

Peter Byrne

Qun Guo

Artem A Oks

Shaoyu Zhou

Rajeev Bashyam Rajan

Mahesh K Sreenivas

Samuel H Smith

Steven R Schmidt

Michael Habben

Eric Ray Christensen

Xiaowei Jiang

Chunjia Li

Sethu M Kalavakur

Gayathri Venkataraman

Avraham Levy

Architecture that addresses an end-to-end solution for logical transactional replication from a shared-nothing clustered database management system, which uses adaptive cloning for high availability. This can be time based using a global logical timestamp. The disclosed architecture, used for refreshing stale clones, does not preserve user transaction boundaries, which is a more complex situation than where the boundaries are preserved. In such a scenario it is probable that for a given data segment no clone of the segment may contain the complete user transaction history, and hence, the history has to be pieced together from the logs of multiple different clones. This is accomplished such that log harvesting is coordinated with the clone state transitions to ensure the correctness of logical replication.

Logical replication in clustered database system with adaptive cloning

Methods, systems, and computer-readable media are disclosed for transitioning clones and clone data access maps in response to node or media failure without blocking queries. In a system, a data request interface is configured to receive a query to access data at a storage device. Data management logic is configured to access clone state information of a plurality of clones including clones in a static state and clones in a transitory state. The data management logic is further configured to process the query to access the data according to the dynamic clone data access map, with functional correctness maintained.

Transitioning clone data maps and synchronizing with a data query

Systems and methods are disclosed that facilitate providing page-level database restore functionality upon detection of a corruption event. Updates to a data page in a database can trigger generation of a snapshot of the data page, and an update log can be maintained that stores information related to page updates. Subsequent snapshots can be generated at predetermined intervals and can trigger truncation of a log segment and initiation of a new log segment. Upon detection of page corruption, a most-recent uncorrupt snapshot of the corrupt page can be identified, copied to the location of the corrupt page in the database, and modified according to the log segment associated with the uncorrupt snapshot to make the page current as of the corrupting event, all of which can be performed to restore the database without having to take the database offline.

Page recovery using volume snapshots and logs

A 'lazy commit' allows database transactions to be committed faster by not waiting for log records of the transaction to be written to disk. When a system crash occurs during the commit process, transactions may be lost because the log data is not available to recreate the transaction. While lazy commits speed up processing, they also increase the potential for data inconsistency. This invention introduces the concept of 'durable reads'&#x2014;transactions that require isolation from lazy commit transactions and which must be guaranteed to read only durably committed data. When durable read transaction attempts to read data changed by a lazy commit transaction, the system ensures the lazy commit transaction's changes are first made durable.

Systems and methods for reading only durably committed data in a system that otherwise permits lazy commit of transactions

To ensure that logs representative of data changes are durably written, localized storage media cache is flushed to force a write to stable storage. Log sequence number tracking is leveraged to determine if log data in a localized storage media cache is likely to have not been written to durable storage. When this condition is likely, a FLUSH_CACHE command is issued to a storage media controller to force a durable write. This allows recovery of data changes to a transactional system even when its associated storage media does not provide write-ordering guarantees. Moreover, flushing of the storage media cache can be accomplished at a frequency that provides an upper bound on a maximum time between a transaction commit and data being made durable.

Enforced transaction system recoverability on media without write-through

A system and method for serialization and/or de-serialization of file system item(s) and associated entity(ies)is provided. A file system 'item' comprises a core class which can include property(ies). Through serialization, a consistent copy of the item and associated entity(ies), such as fragment(s), link(s) with other item(s) and/or extension(s),if any, can be captured. The serialization system includes an identification component that identities entity(ies) associated with an item and a serialization component that serializes the item and associated entity(ies). The serialization component can further serialize a header that includes information associated with the item and associated entity(ies). The header can facilitate random access to the item and associated entity(ies). The serialization system can expose application program interface(s) (API's) that facilitate the copying, moving and/or transfer of an item and its associated entity(ies) from one location to another location.

Serialization of file system item(s) and associated entity(ies)

The subject invention relates to systems and/or methods that perform an authoritative and/or a non-authoritative restore of items in a data store. Backups are provided that include data associated with items and metadata related to the items that can include a history of operations and previous synchronizations to enable replications to converge after restoration. The operations that can be performed on the items include, for example, create, delete, update, move, etc. The items in the backups can be restored along with the metadata, which can be utilized to enable synchronization partners to converge based on a desired type of restoration.

Authoritative and non-authoritative restore

The subject invention relates to systems and methods that enable pausing and/or resuming a backup of file system items. The items are backed up in an order based on a unique property associated with each of the items. When the backup is paused, the unique property associated with a last backed up item prior to the pause is stored. Additionally, a timestamp associated with a snapshot from which the backup was taken can be stored with the unique property. Upon being resumed, the backup can continue from the paused location via utilizing the stored unique property. The backup can be resumed either employing an original snapshot associated with a time prior to the pause or a second snapshot related to a subsequent time. The subject invention can perform an initial backup of file system items and/or an incremental backup of such items. Additionally, the subject invention can support backing up a file system at disparate levels of granularity such as, for example, an item level.

Pausable backups of file system items

A dataset is divided into pieces and stored at multiple locations and the system dynamically increases or decreases the number of storage locations where the pieces of the data set may be stored. A data structure comprises a first data field including a first index and a first element, and one or more data fields each with an index and an element. The elements of the one or more data fields include a token representing a location associated with the index of the first data field. A data row of a data set is mapped to the index of a data field using a second index. The second index is derived from a data row of the data set using a hash function. The second index is then mapped to the index of a data field included in the one or more data fields using a modulus function.

Dynamic fragment mapping

Growing community of inventors

Woodinville, WA, United States of America

Vishal Kathuria