Abstract: A resource manager arranges the resources in a computer system into one or more resource pools. The resource manager allocates a number of active resources and a number of backup resources to a particular resource pool. For each resource managed by the resource manager, the resource manager acquires information that describes the capacity and reliability of the resource. Capacity and reliability information for the particular resource pool is determined based on the capacity and reliability information associated with the resources assigned to the pool. In response to a request, the resource manager may provide an application with resources from several resource pools. The likelihood that the resource manager will be able to provide sufficient resources to the application may be determined based at least in part on the reliability information associated with the several resource pools.

Abstract: A pipeline management service improves the operation of a processing pipeline by monitoring the operation of one or more journal services within the processing pipeline, and adjusting operational parameters of the one or more journal services. In various examples, the pipeline management service determines a model of the processing pipeline by identifying the producers and consumers associated with each of the one or more journal services, and determines a processing flow through the one or more journal services based at least in part on the identities of the associated producers and consumers. The pipeline management service acquires backlog information from each of the one or more journal services, and may adjust the operational parameters of a journal service associated with an identified processing bottleneck. In some examples, the pipeline management service adjusts capacity and reliability parameters of the one or more journal services to improve the operation of the processing pipeline.

Abstract: Methods, systems, and computer-readable media for implementing freshness-sensitive message delivery are disclosed. A stream of events including an earlier event and a later event is received. A first attempt is performed to deliver a message indicative of the earlier event. The first attempt to deliver the message is unsuccessful. One or more additional attempts are scheduled to deliver the message indicative of the earlier event. The additional attempts are scheduled over a period of time following the first attempt. An attempt is performed to deliver an additional message indicative of the later event. The attempt to deliver the additional message is not delayed by the additional attempts to deliver the message.

Abstract: A replication chain comprising one or more replication nodes of a multi-tenant stream management system is assigned to store data records of a partition of a particular data stream. A data record of the partition is received at a selected replication node of the replication chain. In a sequential order, a respective replica of the data record is stored at each replication node of the chain. An acknowledgement of a successful storage of the data record is provided after the replications are completed.

Abstract: A replication chain comprising one or more replication nodes of a multi-tenant stream management system is assigned to store data records of a partition of a particular data stream. A data record of the partition is received at a selected replication node of the replication chain. In a sequential order, a respective replica of the data record is stored at each replication node of the chain. An acknowledgement of a successful storage of the data record is provided after the replications are completed.

Abstract: Methods, systems, and computer-readable media for implementing freshness-sensitive message delivery are disclosed. A stream of events including an earlier event and a later event is received. A first attempt is performed to deliver a message indicative of the earlier event. The first attempt to deliver the message is unsuccessful. One or more additional attempts are scheduled to deliver the message indicative of the earlier event. The additional attempts are scheduled over a period of time following the first attempt. An attempt is performed to deliver an additional message indicative of the later event. The attempt to deliver the additional message is not delayed by the additional attempts to deliver the message.

Abstract: A machine generated service cache that utilizes one or more machine learning classifiers is trained using service requests directed to a human-generated service and service responses generated by the human-generated service in response to the service requests. Once the machine generated service cache has been trained to a predetermined level of performance, the machine generated service cache can be utilized to process actual service requests directed to the human-generated service. The machine generated service cache might be utilized to process service requests for which it is not essential that the returned service response be identical to a response that would be generated by the human-generated service.