Cloudflare, Inc.

Google Inc.

Kenton Taylor Varda

Kyle Kloepper

Alex Dwane Robinson

Dane Orion Knecht

John Graham-Cumming

Marek Przemyslaw Majkowski

Dani Grant

Zachary Aaron Bloom

Ingvar Stepanyan

Brett Joseph Hoerner

Loren Cody Koeninger

Gregory Richard McKeon

Alyson Cabral

Michael Hart

Patrick A Meenan

Umesh Shankar

Darrell Kindred

Lucas Pardue

Glen Patrick Maddern

Vy Nuthuy Ton

Matthew Silverlock

Joshua Tyler Howard

Justin Mazzola Paluska

An object worker is instantiated at a compute server of a distributed cloud computing network, where the object worker includes a single instantiation of a piece of code that solely controls reading/writing to an object. An external communication policy is associated with the first object worker. If the external communication policy does not allow the object worker to send communications with the object to an asset that is external to the distributed cloud computing network, the communication is prevented from being sent. If the external communication policy allows the object worker to send communications with the object to the asset that is external to the distributed cloud computing network, the communication is sent from the first object worker to the asset.

State management and storage with policy enforcement in a distributed cloud computing network

A request is received from a client device at a first datacenter a distributed cloud computing network. The distributed cloud computing network includes multiple datacenters. The received request triggers execution of code at the distributed cloud computing network. The code includes a first function and a second function. A determination is made to execute the first function at the first datacenter and to execute the second function at a second datacenter of the distributed cloud computing network. The first function is executed at the first datacenter to get a first result. The first datacenter causes the second function to be executed at the second datacenter. The first datacenter receives, from the second datacenter, a second result from the execution of the second function. The first datacenter transmits a response to the client device that is based at least in part on the first result and the second result.

Dynamic selection of where to execute application code in a distributed cloud computing network

A first compute server of a distributed cloud computing network receives a request from a first client device for an object to be handled by an object worker that includes a single instantiation of a piece of code that solely controls reading and writing access to the first object. A determination is made that the object worker is instantiated for the object and is currently running in the first compute server, and the piece of code processes the first request. The first compute server receives a message to be processed by the first object worker from a second compute server. The message includes a second request for the object from a second client device connected to the second compute server. The piece of code processes the message and transmits a reply to the second compute server.

State management and object storage in a distributed cloud computing network

A compute server receives a request that triggers execution of a code piece out of multiple code pieces. A single process at the compute server executes the code piece, which is run in an isolated execution environment. Each other code piece runs in other isolated execution environments respectively and executed by the single process. The code piece, when executed, modifies a response to the request. The response is generated based at least in part on the executed code piece. The generated response is transmitted.

Cloud computing platform that executes code in a distributed cloud computing network

Sequential consistency across a distributed cloud computing network is described. A database includes a primary database and multiple read replica databases. Write queries are transmitted to the primary database, and commit tokens are provided to the read replica databases and the clients. Commit tokens are included in requests from clients. If a request for a read operation received at a read replica database does not include a token that is later than a commit token of the most recent update to the read replica database, the read replica database performs the read operation. If a request for a read operation received at a read replica database includes a token that is later than a commit token of the most recent update to the read replica database, the read replica database delays servicing the read update until it receives an update from the primary database with an updated commit token.

Sequential consistency across a distributed cloud computing network

A first compute server of a distributed cloud computing network executes an application that controls reading and writing access to associated persistent data. The first compute server performs a write operation to the persistent data on local storage, notifies a piece of code that controls outgoing messages from the application that the write operation is pending, and transmits write information for the write operation to a set of other compute servers. If an acknowledgement of the write information is received from a quorum of the other compute servers, the application notifies the piece of code that the write operation is confirmed. Periodically the write information is transmitted to an external storage system. If a confirmation that the write information has been written is received from the storage system, the first compute server transmits a write confirmation notice to the other compute servers, which can then delete the write information.

State management and persistent data storage in a distributed cloud computing network

A compute server of a distributed cloud computing network receives a request for an object that is to be handled by an object worker, where the object worker includes a single instantiation of a piece of code that solely controls reading/writing to the object. The object worker is instantiated at the compute server. The compute server enforces an access policy to determine whether the request is allowed to be processed by the object worker. If the request is allowed to be processed by the object worker, the object worker processes the request. If the request is not allowed to be processed by the object worker, the request is blocked.

Concurrency control in an asynchronous event-loop based program environment is described. A program is implemented with an asynchronous event-loop. A piece of code controls events into the program by preventing events from being delivered to the program while a storage operation is executing except for storage completion events. Those events are prevented from being delivered to the program until the storage operation completes and the program is not executing code.

Concurrency control in an asynchronous event-loop based program environment

A compute server receives a first request from a client device that triggers execution of a first third-party code piece. The first request is directed to a first zone. A single process at the compute server executes the first third-party code piece. As a result of executing the first third-party code piece, a second request is generated that triggers execution of a second third-party code piece. The second request is directed to a second zone. The single process executes the second third-party code piece. A response is generated to the first request based at least in part on the executed first third-party code piece and the executed second third-party code piece. The generated response is transmitted to the client device.

Cloud computing platform that executes third-party code in a distributed cloud computing network

A request is received from a client device at a first datacenter of a distributed cloud computing network. The first request triggers execution of code at the distributed cloud computing network. The execution of the code includes transmitting additional requests to destination(s) external to the distributed cloud computing network. A second datacenter of the distributed cloud computing network is selected to execute the code, where the selection is based on an optimization goal. The code is executed at the second datacenter. The first datacenter receives a result from the code being executed at the second datacenter. The first datacenter transmits a response to the client device that is based at least in part on the result.

A compute server of a distributed cloud computing network receives an indication of a hostname in a Transport Layer Security (TLS) handshake message from a client device. The compute server determines that an on-demand code piece is configured to execute on the compute server responsive to receiving an HTTP request for the hostname. Prior to receiving the request for the hostname, the compute server preloads the on-demand code piece for execution. The compute server receives an HTTP request from the client device that triggers execution of the preloaded on-demand code piece. The compute server executed the preloaded on-demand code piece and generates a response to the HTTP request based in part on the executed code piece. The compute server transmits the response to the client device.

Preloading on-demand code pieces in a distributed cloud computing network

A compute server receives a request from a client device that triggers execution of a third-party code piece. The compute server is one of multiple compute servers that are part of a distributed cloud computing network. The request may be an HTTP request and directed to a zone. A single process at the compute server executes the third-party code piece in an isolated execution environment. The single process is also executing other third-party code pieces in other isolated execution environments respectively. A response is generated to the request based at least in part on the executed third-party code piece, and the generated response is transmitted to the client device.

Concurrency control in an asynchronous event-loop based program environment is described. A program is implemented with an asynchronous event-loop. A first piece of code controls events into the program by preventing events from being delivered to the program while a storage operation is executing except for storage completion events. Those events are prevented from being delivered to the program until the storage operation completes and the program is not executing code. A second piece of code controls outgoing messages from the program by preventing outgoing messages from being transmitted when a storage write operation is in progress except for outgoing messages that are storage write operations. Upon the successful completion of the storage write operation, the outgoing messages that were prevented from being transmitted are transmitted.

Growing community of inventors

Palo Alto, CA, United States of America

Kenton Taylor Varda