Abstract: Index management for non-relational database systems may be automatically performed. Performance of queries to a non-relational database may be evaluated to determine whether to create or remove an additional index. An additional index may be automatically created to store a subset of data projected from the non-relational database to utilize when performing a query to the non-relational database instead of accessing data in the non-relational database.

Abstract: A system that implements a data storage service may store data for a database table in multiple replicated partitions on respective storage nodes. In response to a request to back up a table, the service may back up individual partitions of the table to a remote storage system independently and (in some cases) in parallel, and may update (or create) and store metadata about the table and its partitions on storage nodes of the data storage service and/or in the remote storage system. Backing up each partition may include exporting it from the database in which the table is stored, packaging and compressing the exported partition for upload, and uploading the exported, packaged, and compressed partition to the remote storage system. The remote storage system may be a key-value durable storage system in which each backed-up partition is accessible using its partition identifier as the key.

Abstract: Columns of a non-relational data store may be efficiently dropped. A drop column request for a table stored in a non-relational database may be received. In response to the drop column request, schema information for the table may be updated to filter out the column from subsequent read responses to read requests. A backup of the table is created and evaluated to identify those items in the table that include the column. The column is then deleted from the identified items.

Abstract: A distributed database encrypts tables using table encryption keys protected by a client master encryption key. The client may revoke and subsequently restore authorization to access the client master encryption key. A sweeper process of the distributed database examines encrypted tables and identifies changes to the status of a corresponding client master encryption key. A response to an identified change in status is initiated.

Abstract: Changes made to a database table are accumulated, in durable storage, and snapshots of partitions of the table are obtained. For successive snapshots of a partition, the system accesses a previous snapshot, applies changes from the accumulated changes, and stores the updated snapshot to a durable data store. The accumulated changes and the successive partition snapshots are made available to restore the database to any point in time across a continuum between successive snapshots. Although each partition of the table may have a backup snapshot that was generated at a time different from when other partition snapshots were generated, changes from respective change logs may be selectively log-applied to distinct partitions of a table to generate an on-demand backup of the entire table at common point-in-time across partitions. Point-in-time restores of a table may rely upon a similar process to coalesce partition snapshots that are not aligned in time.

Abstract: Techniques are described for enabling a database service to provide a zonal consistency model for read and write operations involving database tables replicated across a plurality of availability zones of a cloud provider network according to some embodiments. A database service routes zonally consistent operations to request routers and storage nodes that are located in a same availability zone as the client device from which the request originates. For example, in contrast to standard write operations and strongly consistent read operations, which are processed by a leader storage node that may be located in a different availability zone than the originating request, zonally consistent operations can be processed by any storage node and, in particular, by a storage node in a same availability zone as the originating client device.

Abstract: A database system may associate functions with a database table. A request to associate a function with a table in a database system may be received. An association between the table and the function may be created. The function may include parameters that are determined from values within the table which are then invoked by a request to perform the function. The associated function may cause the collection of the values prior to performance of the function.

Abstract: Inferences may be obtained to handle access requests at a non-relational database system. An access request may be received at a non-relational database system. The non-relational database system may determine that the access request uses a machine learning model to complete the access request. The non-relational database system may cause an inference to be generated using data items for the access request as input to the machine learning model. The access request may be completed using the generated inference.

Abstract: Replicas may be added to a multi-leader replica group for a data set. A new replica may be added to a group of replicas by copying a version of the data set according to a point in time to the new replica. As updates may occur at different ones of the replicas when the new replica is added, updates that occurred at a source replica for the snapshot and other replicas may be replicated by the different replicas according to the point in time at which the snapshot is created without blocking updates at any of the replicas in the replica group.

Abstract: Methods and apparatus for conditional master election in a distributed database are described. A plurality of replicas of a database object are stored by a distributed database service. Some types of operations corresponding to client requests directed at the database object are to be coordinated by a master replica. Client access to the database object is enabled prior to election of a master replica. In response to a triggering condition, a particular replica is elected master. The master coordinates implementation of operations with one or more other replicas in response to client requests.

Abstract: Inferences may be obtained to handle access requests at a non-relational database system. An access request may be received at a non-relational database system. The non-relational database system may determine that the access request uses a machine learning model to complete the access request. The non-relational database system may cause an inference to be generated using data items for the access request as input to the machine learning model. The access request may be completed using the generated inference.

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: Methods and apparatus for conditional master election in a distributed database are described. A plurality of replicas of a database object are stored by a distributed database service. Some types of operations corresponding to client requests directed at the database object are to be coordinated by a master replica. Client access to the database object is enabled prior to election of a master replica. In response to a triggering condition, a particular replica is elected master. The master coordinates implementation of operations with one or more other replicas in response to client requests.

Abstract: An interface for requesting, and technique for generation of, a backup of a past state of a database table are provided. Changes made to a database table are accumulated, in durable storage, and snapshots of partitions of the table are obtained. The accumulated changes and the successive partition snapshots are used to generate a past state of the database at any point in time across a continuum between successive snapshots. Although each partition of the table may have a snapshot that was generated at a time different from when other partition snapshots were generated, changes from respective change logs may be selectively log-applied to distinct partitions of a table to generate backup in the past of the entire table at common point-in-time across partitions.

Abstract: A distributed database management system comprises database nodes located in a plurality of availability zones. The system receives a request to store a table, and determines to store a number of replicas of the table in one or more availability zones. The system obtains information indicating that one of the availability zones has been designated as preferred. The system identifies a set of database nodes in the preferred availability zone, where the number of database nodes in the subset is sufficient to determine quorum among the replicas. The replicas are stored on the identified set of database nodes.

Abstract: A distributed database encrypts a table using a table encryption key protected by a client master encryption key. The encrypted table is replicated among a plurality of nodes of the distributed database. The table encryption key is replicated among the plurality of nodes, and is stored on each node in a respective secure memory. In the event of node failure, a copy of the stored key held by another member of the replication group is used to restore a node to operation. The replication group may continue operation in the event of a revocation of authorization to access the client master encryption key.

Abstract: Information describing changes to a collection of items maintained by a database may be stored in a log file. The information in the log file may be converted into a stream of records describing the changes. The records may be directed to a computing node selected for performing a trigger function in response to the change, based on applying a hash function to a portion of the record, identifying a hash space associated with a value output by the hash function, and mapping from the hash space to the selected computing node.

Abstract: An interface for requesting, and technique for generation of, a backup of a past state of a database table are provided. Changes made to a database table are accumulated, in durable storage, and snapshots of partitions of the table are obtained. The accumulated changes and the successive partition snapshots are used to generate a past state of the database at any point in time across a continuum between successive snapshots. Although each partition of the table may have a snapshot that was generated at a time different from when other partition snapshots were generated, changes from respective change logs may be selectively log-applied to distinct partitions of a table to generate backup in the past of the entire table at common point-in-time across partitions.

Abstract: A distributed database encrypts tables using table encryption keys protected by a client master encryption key. The client may revoke authorization to access the client master encryption key. Subsequent to a revocation of authority to access the client master encryption key, the distributed database generates interim snapshots of the table using the table encryption key. Also subsequent to the revocation, the distributed database generates a backup of the table using a backup encryption key protected by the client master encryption key.

Abstract: A transmission format cache may be implemented at a database storage node. Versions of data items stored in a database at the database storage node may be processed according to anticipated access requests to generate an anticipated access responses. The anticipated access responses are then stored in a transmission format cache to provide low latency reads of the data items. The versions of the data items may be processed as a result of updates to the database items. The database storage node may be one of a plurality of database storage nodes implementing a distributed database system with the transmission format cache implementing a portion of a distributed response cache providing low latency, eventually consistent or consistent reads of data items in a distributed database.