Microsoft Technology Licensing, LLC

Steven John Lindell

Sandeep Lingam

Girish Mittur Venkataramanappa

Sameer Arun Verkhedkar

Zhengguo Sun

Varun Kunjbihari Tibrewal

Colin Neville

Sameet Harishanker Agarwal

Srikumar Rangarajan

Jeffrey A East

Goetz Graefe

Jakub Kulesza

Sarvesh Singh

Michael I Zwilling

Hari Krishna Sunder

Rishabh Singh Ahluwalia

Rande Blackman

Albert L Lingelbach

Craig G Zastera

Luis Felipe

Tianhui Shi

Methods for snapshot isolation query transactions in distributed systems are performed by systems and devices. Distributed executions of queries are performed in a processing system according to an isolation level protocol for data management and data versioning across one or more data sets, one or more compute pools, etc., within a logical server via a single transaction manager that oversees the isolation semantics and data versioning. Read transactions of queries are performed lock-free via the isolation semantics, and instant rollbacks, point-in-time queries, single-phase commits in the distributed systems are also provided. Abort and cleanup operations are performed based on a distributed abort protocol and a determined oldest active transaction for the system in which the single transaction manager does not track read-only transactions, and client nodes do not maintain commit tables for transactions.

Snapshot isolation query transactions in distributed systems

Methods for rowgroup consolidation with delta accumulation and versioning in distributed systems are performed. The systems provide performant methods of row storage that enable versioned modifications of data while keeping and allowing access to older versions of the data for point-in-time transactions. The accumulation of valid rows, deletes, and modifications is maintained in blobs for rowgroups until a size threshold is reached, at which point the rows are moved into a columnar compressed form. Changes to data and associated metadata are stored locally and globally via appends, maintaining logical consistency. Metadata is stored in footers of files allowing faster access to the metadata and its associated data for transactions and instant rollback via metadata version flipping for aborted transactions, as well as lock-free reads of data.

Rowgroup consolidation with global delta accumulation and versioning in distributed systems

A Distributed Availability Group (DAG) spans two AGs, each spanning one or more replica nodes and functioning as primary or secondary AG. A primary AG is replicated to the secondary AG synchronously or asynchronously. A failover in the DAG results in the AGs swapping their roles. Multiple DAGs can be linked together as a chain, which provides many useful features including disaster recovery across geographical regions, massive read scale (numerous readable secondary nodes), online migration of databases (across different operating systems and computing environments). The systems using DAGs can replicate databases across multiple independent high availability (HA) failover clusters using complex replication topologies and allow for manual failover and failback. The systems allow chaining of multiple AGs to provision a treelike structure of replicas and numerous secondary replicas without impacting performance. The systems automatically seed new database replicas to facilitate building a complex topology of DAGs.

Distributed availability groups of databases for data centers including seeding, synchronous replications, and failover

Distributed availability groups of databases for data centers for providing massive read scale

Distributed availability groups of databases for data centers including different commit policies

A Distributed Availability Group (DAG) spans two AGs, each spanning one or more replica nodes and functioning as primary or secondary AG. A primary AG is replicated to the secondary AG synchronously or asynchronously. A failover in the DAG results in the AGs swapping their roles. Multiple DAGs can be linked together as a chain, which provides many useful features including disaster recovery across geographical regions, massive read scale (numerous readable secondary nodes), online migration of databases (across different operating systems and computing environments). The systems using DAGs can replicate databases across multiple independent high availability (HA) failover clusters using complex replication topologies and allow for manual failover and failback. The systems allow chaining of multiple AGs to provision a treelike structure of replicas and numerous secondary replicas without impacting performance. The systems automatically seed new database replicas that to facilitate building a complex topology of DAGs.

Distributed availability groups of databases for data centers

Distributed availability groups of databases for data centers including failover to regions in different time zones

A method of invalidating an unused log file in a database includes examining a discard log information storage area and determining if a discard log exists. If the discard log exists, the log is invalidated. The headers of each of the log files that have been fully invalidated are then invalidated. The discard log is then fully invalidated and a new log file may be generated on top of the discarded log. The method may be interrupted by a failure, a changeover, or crash and will be repeated until it completes.

Robust end-of-log processing

An on-line dynamic shrink facility sets an allocation fence in a file and moves data from above the fence to below the fence. The fence is based on a target size for the file. When all the data above the fence is moved below the fence, the space above the fence is freed. The target size can be a specified size, an original size for the file when it was created, or a size calculated by examining the amount of space used and adding in an amount of desired free space. Resources that are required to move the data, such as old and new allocation units for the data, or that will be impacted by movement of the data, such as index entries, are locked or latched to prevent corruption of the data while it is being moved. The operations are performed in discreet transactions which are immediately committed to minimize the impact on other users of the file.

On-line dynamic file shrink facility

A linked data structure verification system to verify the integrity of at least one linked data structure simultaneously by way of a verification setup phase and an integrity verification phase. Individual nodes are retrieved from a memory device and examined seriatim in optimal memory device location order. Nodes are retrieved and examined in optimal memory device location order for maximum memory device retrieval performance. Expected and/or actual node information about nodes in a given linked data structure are temporarily stored as records in an integrity verification table for only as much time as is necessary to verify any part of the node information prior to excising one or more unnecessary records from the integrity verification table.

Linked data structure integrity verification system which verifies actual node information with expected node information stored in a table

Growing community of inventors

Bellevue, WA, United States of America

Steven John Lindell