Synchronous replication in a distributed storage environment

One or more computer storage devices having computer-executable instructions embodied thereon, that when executed by a computing system having a processor and memory, cause the computing system to perform a method for synchronously geo-replicating data in a distributed storage environment, the method comprising:

• receiving data from a client at a primary data store, the primary data store comprising a log and one or more memory tables;
• locking an object in the one or more memory tables of the primary data store that is associated with the data at the primary data store;
• from the primary data store, communicating the data to a secondary data store with a record, the record including information permitting a log at the secondary data store to commit the data at the secondary data store, the received data not being accessible from the one or more memory tables of the primary data store prior to being committed;
• writing the data to the log at the secondary data store;
• receiving a first acknowledgment from the secondary data store that the data was written to the log at the secondary data store, the receiving being while the object is locked in the one or more memory tables of the primary data store, the first acknowledgement being associated with the record;
• in response to the receiving the first acknowledgment from the secondary data store, committing the data at the primary data store, the committing the data at the primary data store comprising writing the received data to the one of the memory tables of the primary data store;
• communicating, from the primary data store to the secondary data store, an instruction to commit the data;
• receiving a second acknowledgment from the secondary data store, the second acknowledgment acknowledges that the data was committed at the secondary data store and is thereby rendered accessible to the client at the secondary data store;
• in response to receiving the second acknowledgment at the primary data store, unlocking the locked object;
• allowing access to the received data from the at least one memory table;
• and
• communicating an acknowledgement of committal of the data at the primary data store to the client.

A computer-implemented method in a distributed storage environment utilizing a processor and memory for synchronously geo-replicating data in the distributed storage environment, the method comprising:

• receiving data from a client at a primary data store, the primary data store comprising a log and one or more memory tables, the data having a sequential order;
• communicating from the primary data store to a secondary data store the data and an identifier of the data that is generated within the distributed computing environment to facilitate synchronous committal of data to the secondary data store while the data is inaccessible to the client at the primary data store, the primary data store and the secondary data store being in different geographic regions, the received data not being accessible from the one or more memory tables of the primary data store prior to being committed;
• receiving the data at the secondary data store from the communicating, the data being non-sequentially received data;
• at the secondary data store, writing the non-sequentially received data to a log at the secondary data store, the non-sequentially received data being inaccessible to the client upon completion of the writing;
• communicating, from the secondary data store to the primary data store, an acknowledgement of writing the data to the log at the secondary data store, wherein the acknowledgement includes at least a portion of the identifier;
• in response to receiving the acknowledgement of writing the data to the log at the secondary data store, committing the data at the primary data store to the primary data store thereby rendering the data at the primary data store accessible to the client, the committing of the data at the primary data store to the primary data store comprising writing the data in a sequential manner to at least one of the one of the memory tables of the primary data store prior to being accessible to the client;
• allowing access to the received data from the at least one memory table;
• communicating, from the primary data store to the secondary data store, an instruction to commit the data at the secondary data store, wherein the instruction includes, at least in part, a portion of the identifier;
• and
• committing, in response to the instruction to commit, the non-sequentially received data from the log at the secondary data store to the secondary data store thereby providing the sequential order to the non-sequentially received data and rendering the data at the secondary data store accessible to the client, the committing of the non-sequentially received data from the log at the secondary data store to the secondary data store comprising committing the non-sequentially received data in a sequential manner.