Techniques for deterministically routing database requests to database servers

A non-transitory computer readable storage medium configured to store instructions that, when executed by at least one processor included in a database server, cause the database server to manage a plurality of database engines, by carrying out steps that include:

• concurrently executing the plurality of database engines;
• and
• in response to receiving a request, from a routing server, to perform an input/output operation to a database file of a plurality of database files, wherein the request is provided to the database server based on a hash of a user ID and a count of a group of database servers identified as capable of accessing the database file:
• selecting, among the plurality of database engines, a database engine that is available to perform the input/output operation;
• performing at least one operation to make the database file accessible to the database engine;
• obtaining an exclusive lock on the database file, wherein obtaining the exclusive lock comprises updating metadata associated with the database file in response to determining that the exclusive lock is obtained, and wherein, in response to determining that the exclusive lock is not obtained, the steps further include extracting, from the metadata, information about a different database server that has obtained the exclusive lock on the database file, and then providing the information to the routing server;
• in response to determining that the exclusive lock is obtained, causing the database engine to perform the input/output operation to the database file;
• and
• in response to determining that the exclusive lock is obtained, providing, to the routing server, a response in accordance with the input/output operation.

A database server configured to manage a plurality of database engines, the database server comprising at least one processor configured to cause the database server to carry out steps that include:

• concurrently executing the plurality of database engines;
• and
• in response to receiving a request, from a routing server, to perform an input/output operation to a database file of a plurality of database files, wherein the request is provided to the database server based on a hash of a user ID and a count of a group of database servers identified as capable of accessing the database file:
• selecting, among the plurality of database engines, a database engine that is available to perform the input/output operation;
• performing at least one operation to make the database file accessible to the database engine;
• obtaining an exclusive lock on the database file, wherein obtaining the exclusive lock comprises updating metadata associated with the database file in response to determining that the exclusive lock is obtained, and wherein, in response to determining that the exclusive lock is not obtained, the steps further include extracting, from the metadata, information about a different database server that has obtained the exclusive lock on the database file, and then providing the information to the routing server;
• in response to determining that the exclusive lock is obtained, causing the database engine to perform the input/output operation to the database file;
• and
• in response to determining that the exclusive lock is obtained, providing, to the routing server, a response in accordance with the input/output operation.