Abstract: A content management system stores distributed data tables containing key-value pairs across a plurality of nodes and maintains a plurality of slices, with each slice corresponding to a contiguous key range across the data tables. The content management system may rebalance data among nodes by performing operations such as transferring, merging, or splitting slices. Each operation may be accomplished by performing multiple actions, and each action may cause change in states for slices. During an operation, slices may go through a series of state transitions. For each state transition, the content management system may record a timestamp when the state transition took place and the content management system may maintain a log that records timestamped state transitions associated with slices. The content management system may also perform various invariant checks and determine whether to reject or allow a state transition based on results of invariant checks.

Abstract: Techniques are provided for a linearizable primary-secondary database system that replicates database data with eventual consistency. The techniques include obtaining a logical clock value of a logical clock that orders transactions committed at a primary database system that includes a primary database. The logical clock value is obtained after a write transaction is committed at the primary database system and while processing a read request. The logical clock value is used to determine whether a secondary database is in a consistent replication state corresponding to the logical clock value with respect to the primary database. The read request is maintained in a linger state while the secondary database is not yet in the consistent replication state. The read request exits the linger state and is allowed to be processed against the secondary database after the secondary database has reached the consistent replication state.

Abstract: This invention relates to a system and methods for determining the placement of an object in a distributed key-value store by mapping the object to nodes in multidimensional hyperspace. A search function supports efficient object retrieval, even when the search query requests multiple objects and specifies them through non-primary keys. In response to a search query, the search is translated into hyperregions in the hyperspace to determine the set of nodes that hold the queried data object. The number of contacted nodes and the number of scanned objects are significantly reduced in comparison to prior art techniques.

Abstract: Techniques are provided for a linearizable primary-secondary database system that replicates database data with eventual consistency. The techniques include obtaining a logical clock value of a logical clock that orders transactions committed at a primary database system that includes a primary database. The logical clock value is obtained after a write transaction is committed at the primary database system and while processing a read request. The logical clock value is used to determine whether a secondary database is in a consistent replication state corresponding to the logical clock value with respect to the primary database. The read request is maintained in a linger state while the secondary database is not yet in the consistent replication state. The read request exits the linger state and is allowed to be processed against the secondary database after the secondary database has reached the consistent replication state.

Abstract: Shared hash map implementations provide a linearizable programmatic interface to calling threads and support lock-free hash map operations. In addition to supporting lock-free operation, the shared hash map implementations are especially useful where the hash map is stored in a fixed-sized memory location such as, for example, in a shared memory segment, or in other situations where resizing the hash map is impractical or not desired.

Abstract: This invention relates to a system and methods for determining the placement of an object in a distributed key-value store by mapping the object to nodes in multidimensional hyperspace. A search function supports efficient object retrieval, even when the search query requests multiple objects and specifies them through non-primary keys. In response to a search query, the search is translated into hyperregions in the hyperspace to determine the set of nodes that hold the queried data object. The number of contacted nodes and the number of scanned objects are significantly reduced in comparison to prior art techniques.

Abstract: This invention relates to a system and methods for determining the placement of an object in a distributed key-value store by mapping the object to nodes in multidimensional hyperspace. A search function supports efficient object retrieval, even when the search query requests multiple objects and specifies them through non-primary keys. In response to a search query, the search is translated into hyperregions in the hyperspace to determine the set of nodes that hold the queried data object. The number of contacted nodes and the number of scanned objects are significantly reduced in comparison to prior art techniques.

Abstract: An efficient fault-tolerant event ordering service as well as a simplified approach to transaction processing based on global event ordering determines the order of interdependent operations in a distributed system. The fault-tolerant event ordering service externalizes the task of tracking dependencies to capture a global view of dependencies between a set of distributed operations in a distributed system. A novel protocol referred to as linear transactions coordinates distributed transactions with Atomicity, Consistency, Isolation, Durability (ACID) semantics on top of a sharded data store. The linear transactions protocol achieves scalability by distributing the coordination task to only those servers that hold relevant data for each transaction and achieves high performance by serializing only those transactions whose concurrent execution could potentially yield a violation of ACID semantics.