Method for applying dynamic data block caching automation for high-speed data access based on computational storage

A database (DB) system comprising:

• a computational storage in which a DB is established;
• and
• a database management system (DBMS) comprising one or more processors configured to synchronize an external computational cache (ECC) and an internal computational cache (ICC) of the computational storage, to generate an offloading execution code that defines operation information necessary for query computation offloading based on a query requested by a client, and to process the offloading execution code by using the ECC and the ICC which are synchronized,
• wherein the one or more processors of the DBMS are further configured to:
• process the offloading execution code by distinguishing by at least one of ECC scanning, ICC scanning, computational storage driver (CSD) scanning by using the ECC and the ICC which are synchronized,
• wherein the processing comprises:
• performing a first identification of identifying, by the DMBS, whether data exists in the ECC;
• performing a second identification of, when it is identified that the data exists in the ECC at the first identification, identifying, by the DBMS, whether all of the data that is recorded on the offloading execution code and needs to be scanned exist in the ECC;
• and
• when it is identified that all of the data exist in the ECC at the second identification, processing, by the DBMS, the offloading execution code in the ECC, and
• wherein the one or more processors of the DBMS are further configured to conditionally refrain from transmitting the offloading execution code to the computational storage when the ECC contains all data required for execution.

A query execution method comprising:

• generating, by a database management system (DMBS), an offloading execution code that defines operation information necessary for query computation offloading based on a query requested by a client;
• and
• processing, by the DBMS, the offloading execution code by using an external computational cache (ECC) which is a cache of the DBMS,
• wherein the method further comprises:
• processing the offloading execution code by distinguishing by at least one of ECC scanning, ICC scanning, computational storage driver (CSD) scanning by using the ECC and the ICC which are synchronized,
• wherein the processing comprises:
• performing a first identification of identifying, by the DMBS, whether data exists in the ECC;
• performing a second identification of, when it is identified that the data exists in the ECC at the first identification, identifying, by the DBMS, whether all of the data that is recorded on the offloading execution code and needs to be scanned exist in the ECC;
• and
• when it is identified that all of the data exist in the ECC at the second identification, processing, by the DBMS, the offloading execution code in the ECC, and
• wherein the method further comprises, when it is identified that all data required by the offloading execution code exists in the ECC, refraining, by the DBMS, from delivering the offloading execution code to the computational storage.