Abstract: A database search method combines SQL command, complex event processing, and keyword search. Event data records may be collected based on the events specified in a query. A complex event processing analysis is used to determine the records that meet the correct sequence of events. The keyword search is used to filter user data records that have the attributes specified in the query. A SQL command may be used to further analyze or group the data records found. One or more metrics of the filtered results are presented as the result of the query. To run a new query efficiently, user identifiers are distributed using hashes. Adaptive sampling is used to determine whether the search result is within a desired margin of error. An efficient search for a specific condition is performed using a pace computation that jumps into a location of user data records that have the attributes.

Abstract: Techniques are described for asynchronous communication of transaction data to client processes. A transaction request may be sent from a client process to a contract engine, which may perform multiple service calls to network services to generate transaction data for the transaction. The transaction data may be stored in an aggregate contract data object, and provided asynchronously to the client process for presentation via a user interface such as an e-commerce application. In some cases, the client process may, subsequently to the initial transaction request, send one or more requests for portions of the transaction data. Such requests may be sent asynchronously with respect to the initial transaction request. In some cases, the client process may provide callback information specifying a network location to which transaction data may be asynchronously communicated after it has been generated.

Abstract: Techniques are described for parallel processing of items, specified for purchase in a transaction. A transaction request may be received, specifying a plurality of items in a transaction. Multiple subsets of one or more items may be determined that are suitable for parallel processing by multiple instances of a network services, the subsets determined based on configuration information describing a parallel processing capability of the network service specified during registration. Multiple service calls may be made to cause the multiple instances of the network service to generate transaction data corresponding to the subsets. The transaction data may be incorporated into an aggregate contract data object on receipt, and provided to client process(es) for presentation via a user interface such as an e-commerce application.

Abstract: A data aggregation service is configured to minimize the number of service calls made to network services. The data aggregation service might be configured to cache objects returned in response to service calls to network services in a shared data object. A hash of the input values in the service calls to the network services may be utilized to determine whether an object referenced by the cache can be utilized instead of making another service call. The data aggregation service might also be configured to utilize change tracking to determine when calls are to be made to dependent services. The data aggregation service might also be configured to utilize data-specific time to live (“TTL”) values, to allow network services to specify the data they are interested in at a granular level, and/or to perform automated optimization of TTL values. Other optimizations might also be implemented by the data aggregation service.