Abstract: Techniques for a context preservation during server restart are described. A restart of a database server having at least one open communications session with a database client is initiated. The database server can be executed by a compute instance hosted by a computer system of a provider network. A restarted database server is launched. The restarted database server obtains session state data to resume the at least one open communications session with the database client, the session state data including one or more session variables. The restarted database server receives a transaction from the database client.

Abstract: A database associated with a primary database cluster may be replicated in a backup database cluster located in a different location in order to provide a highly-available fault tolerant database service. The databases are reconciled through a cluster management module distributed in each database cluster. The cluster management module uses a set of reconciliation data structures to track locally the reconciled states of each database in each database cluster, the operations made locally to the databases in each database cluster, and the author of the operations. The cluster management module in each database cluster engages in a stateless messaging protocol using the set of reconciliation data structures to determine whether or not the databases may be reconciled.

Abstract: Aspects of the subject matter described herein relate to a buffer pool for a database system. In aspects, secondary memory such as solid state storage is used to extend the buffer pool of a database system. Thresholds such as hot, warm, and cold for classifying pages based on access history of the pages may be determined via a sampling algorithm. When a database system needs to free space in a buffer pool in main memory, a page may be evicted to the buffer pool in secondary memory or other storage based on how the page is classified and conditions of the secondary memory or other storage.

Abstract: A database associated with a primary database cluster may be replicated in a backup database cluster located in a different location in order to provide a highly-available fault tolerant database service. The databases are reconciled through a cluster management module distributed in each database cluster. The cluster management module uses a set of reconciliation data structures to track locally the reconciled states of each database in each database cluster, the operations made locally to the databases in each database cluster, and the author of the operations. The cluster management module in each database cluster engages in a stateless messaging protocol using the set of reconciliation data structures to determine whether or not the databases may be reconciled.

Abstract: Aspects of the subject matter described herein relate to a buffer pool for a database system. In aspects, secondary memory such as solid state storage is used to extend the buffer pool of a database system. Thresholds such as hot, warm, and cold for classifying pages based on access history of the pages may be determined via a sampling algorithm. When a database system needs to free space in a buffer pool in main memory, a page may be evicted to the buffer pool in secondary memory or other storage based on how the page is classified and conditions of the secondary memory or other storage.

Abstract: Aspects of the subject matter described herein relate to a buffer pool for a database system. In aspects, secondary memory such as solid state storage is used to extend the buffer pool of a database system. Thresholds such as hot, warm, and cold for classifying pages based on access history of the pages may be determined via a sampling algorithm. When a database system needs to free space in a buffer pool in main memory, a page may be evicted to the buffer pool in secondary memory or other storage based on how the page is classified and conditions of the secondary memory or other storage.

Abstract: Embodiments are directed to providing distributed database service upgrades of database server instances in a computer cluster using multiple database server instances and to monitoring and maintaining a distributed database service during upgrade. In an embodiment, each computer system in a computer cluster instantiates at least two different database server instances on each of the nodes in the cluster. The first database server instances are configured to operate using a current distributed database version and the second instances are configured to operate using a new, updated distributed database service version. The computer system receives an indication that the distributed database service is to be upgraded. Then, based on the received indication, the computer system migrates database replicas from the first database server instances to the second database server instances which operate the new, updated service version, substantially without user-visible downtime.

Abstract: Embodiments are directed to providing distributed database service upgrades of database server instances in a computer cluster using multiple database server instances and to monitoring and maintaining a distributed database service during upgrade. In an embodiment, each computer system in a computer cluster instantiates at least two different database server instances on each of the nodes in the cluster. The first database server instances are configured to operate using a current distributed database version and the second instances are configured to operate using a new, updated distributed database service version. The computer system receives an indication that the distributed database service is to be upgraded. Then, based on the received indication, the computer system migrates database replicas from the first database server instances to the second database server instances which operate the new, updated service version, substantially without user-visible downtime.

Abstract: Methods, systems, and computer-readable media are disclosed for partitioning contended synchronization objects. A particular method determines a contention-free value of a performance metric associated with a synchronization object of a data structure. A contended value of the performance metric is measured, and the synchronization object is partitioned when the contended value of the performance metric exceeds a multiple of the contention-free value of the performance metric.

Abstract: Systems and methods for scheduling computing device resources include a scheduler that maintains multiple queues. Requests are placed in one of the multiple queues depending on how much resource time the requests are to receive and when they are to receive it. The queue that a request is placed into depends on a pool bandwidth defined for a pool that includes the request and a bandwidth request. A request has an importance associated therewith that is taken into account in the scheduling process. The scheduler proceeds through the queues in a sequential and circular fashion, taking a work item from a queue for processing when that queue is accessed.

Abstract: A database can have multiple requests applied at one time. Each of these requests requires a specific amount of server resources. There can be a differentiation of user-submitted workloads between each other. These workloads are a set of queries submitted by different users. Each query can have specific resource limits. In addition, each set can have specific resource limits.

Abstract: Aspects of the subject matter described herein relate to a buffer pool for a database system. In aspects, secondary memory such as solid state storage is used to extend the buffer pool of a database system. Thresholds such as hot, warm, and cold for classifying pages based on access history of the pages may be determined via a sampling algorithm. When a database system needs to free space in a buffer pool in main memory, a page may be evicted to the buffer pool in secondary memory or other storage based on how the page is classified and conditions of the secondary memory or other storage.

Abstract: A database communication system is described herein that structures communications in a way that provides lower overhead tracking, statistics, semantics for closing a communication, and reliability. The system provides communication namespaces that organize communications by component, purpose, and instance, which allow database servers to implicitly create communication-related objects without central coordination. The database communication system enables group-based communications that streamline the development of complex distributed components and protocols by providing creation and management of communications namespaces, centralized cleanup support, and centralized monitoring. These features allow the system to be highly distributed, with no one single coordinator of operations, and still provide reliable communications. Thus, the system allows databases to be spread across multiple servers while keeping the burden on database server developers of managing communications between the servers low.

Abstract: Aspects of the subject matter described herein relate to deadlock detection in distributed environments. In aspects, nodes that are part of the environment each independently create a local wait-for graph. Each node transforms its local wait-for graph to remove non-global transactions that do not need resources from multiple nodes. Each node then sends its transformed local wait-for graph to a global deadlock monitor. The global deadlock monitor combines the local wait-for graphs into a global wait-for graph. Phantom deadlocks are detected and removed from the global wait-for graph. The global deadlock monitor may then detect and resolve deadlocks that involve global transactions.

Abstract: Systems and methods for scheduling computing device resources include a scheduler that maintains multiple queues. Requests are placed in one of the multiple queues depending on how much resource time the requests are to receive and when they are to receive it. The queue that a request is placed into depends on a pool bandwidth defined for a pool that includes the request and a bandwidth request. A request has an importance associated therewith that is taken into account in the scheduling process. The scheduler proceeds through the queues in a sequential and circular fashion, taking a work item from a queue for processing when that queue is accessed.

Abstract: A database can have multiple requests applied at one time. Each of these requests requires a specific amount of server resources. There can be a differentiation of user-submitted workloads between each other. These workloads are a set of queries submitted by different users. Each query can have specific resource limits. In addition, each set can have specific resource limits.