To distribute messages to subscribers that are located in multiple regions, a data center will receive messages that to be published to the subscribers. The data center will then perform a limited replication process to other data centers in other regions by accessing an interest map indicating additional data centers at which subscribers have recently expressed interest. The data center will transmit the messages to only that group of additional data centers for replication but not to other data centers at which no interest has been recently expressed.

To distribute messages to subscribers that are located in multiple regions, a data center will receive messages that to be published to the subscribers. The data center will then perform a limited replication process to other data centers in other regions by accessing an interest map indicating additional data centers at which subscribers have recently expressed interest. The data center will transmit the messages to only that group of additional data centers for replication but not to other data centers at which no interest has been recently expressed.

According to an aspect, a computer-implemented method for performing distributed communication operations includes receiving a request, by a first computing system, to perform a distributed communication operation and obtaining, by the first computing system, a tree structure for performing the distributed communication operation, wherein the first computing system is a root node of the tree structure. The method also includes creating, by the first computing system, a message having header information and a payload for the distributed communication operation and transmitting, by the first computing system, a portion of the message to each child node of the first computing system, wherein the portion transmitted to each child node is unique.

According to an aspect, a computer-implemented method for performing distributed communication operations includes receiving a request, by a first computing system, to perform a distributed communication operation and obtaining, by the first computing system, a tree structure for performing the distributed communication operation, wherein the first computing system is a root node of the tree structure. The method also includes creating, by the first computing system, a message having header information and a payload for the distributed communication operation and transmitting, by the first computing system, a portion of the message to each child node of the first computing system, wherein the portion transmitted to each child node is unique.

Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a system that controls content distribution using a feedback loop. Content is distributed over multiple different online channels using a same initial selection value for distribution over each different online channel. An observed user actions required for distribution of the content over the multiple different online channels is received through a feedback loop and for multiple different distributions of the content. Based on the observed user actions received through the feedback loop, a predicted user action rate is determined for the multiple different distributions across the multiple different online channels. The selection value is adjusted based on a difference between the predicted user action rate and a reference distribution amount specified by a provider of the content. The content is distributed over the multiple different online channels using the adjusted selection value.

Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a system that controls content distribution using a feedback loop. Content is distributed over multiple different online channels using a same initial selection value for distribution over each different online channel. An observed user actions required for distribution of the content over the multiple different online channels is received through a feedback loop and for multiple different distributions of the content. Based on the observed user actions received through the feedback loop, a predicted user action rate is determined for the multiple different distributions across the multiple different online channels. The selection value is adjusted based on a difference between the predicted user action rate and a reference distribution amount specified by a provider of the content. The content is distributed over the multiple different online channels using the adjusted selection value.

Systems and methods for reducing data movement in a computer system. The systems and methods use information or knowledge about the structure of an algorithm, operations to be executed at a receiving processing unit, variables or subsets or groups of variables in a distributed algorithm, or other forms of contextual information, for reducing the number of bits transmitted from at least one transmitting processing unit to at least one receiving processing unit or storage device.

Systems and methods for reducing data movement in a computer system. The systems and methods use information or knowledge about the structure of an algorithm, operations to be executed at a receiving processing unit, variables or subsets or groups of variables in a distributed algorithm, or other forms of contextual information, for reducing the number of bits transmitted from at least one transmitting processing unit to at least one receiving processing unit or storage device.

A system for managing access to a shared resource includes a plurality of nodes is described. The system may determine a quota for each of a plurality of nodes in view of one or more factors associated with a respective node and set a local counter stored in a memory of each of the nodes to the value of the relevant quota. In response to a request to access a resource made by an end device, the system may serve the request with a first node in the nodes, wherein the local counter stored in the memory of the first node is more than zero, and decrement the local counter stored in the memory of the first node by one after serving the request. Synchronization data is sent from each of the nodes, including the value of the local counter stored in the memory of the relevant node.

A system for managing access to a shared resource includes a plurality of nodes is described. The system may determine a quota for each of a plurality of nodes in view of one or more factors associated with a respective node and set a local counter stored in a memory of each of the nodes to the value of the relevant quota. In response to a request to access a resource made by an end device, the system may serve the request with a first node in the nodes, wherein the local counter stored in the memory of the first node is more than zero, and decrement the local counter stored in the memory of the first node by one after serving the request. Synchronization data is sent from each of the nodes, including the value of the local counter stored in the memory of the relevant node.