Selecting between hydration-based scanning and stateless scale-out scanning to improve query performance

Patent metadata

Field	Value
Publication number	US-11593367-B1
Application number	US-202117489532-A
Country	US
Kind code	B1
Filing date	Sep 29, 2021
Priority date	Sep 29, 2021
Publication date	Feb 28, 2023
Grant date	Feb 28, 2023

Abstract

When a query is received by a stateful data processing service, the service determines, for each table scan (and associated operations) of a query, whether to select the table scan for execution by a stateless data processing service. The selected table scans are sent to the stateless data processing service for execution, and results are received by the stateful data processing service. The stateful data processing service may also execute other table scans of the query locally, against a local data cache. If the data is not present in the local data cache, then the stateful data processing service will copy the table data into the local data cache before executing the table scan. A query result based on the remote and/or local table scans may then be returned to the client.

First claim

What is claimed is:

1. A system, comprising:
one or more processors; and
one or more memories, wherein the one or more memories have stored thereon instructions, which when executed by the one or more processors of a provider network, cause the one or more processors to implement a stateful data processing service for a plurality of clients, wherein the stateful data processing service is configured to, for a given client of the plurality of clients:
receive, from the client, a query for a database, wherein the query indicates at least a plurality of table scans and associated operations to be performed on a plurality of tables of the database;
select, based on one or more criteria, one or more table scans and associated operations from among the plurality of table scans and associated operations;
send an indication of the one or more table scans and associated operations to a stateless data processing service; and

one or more other processors; and
one or more other memories, wherein the one or more other memories have stored thereon instructions, which when executed by the one or more other processors of a provider network, cause the one or more other processors to implement a stateless data processing service for the plurality of clients, wherein the stateless data processing service is configured to, for the given client:
receive, from the stateful data processing service, the indication of one or more table scans and associated operations;
perform the one or more table scans and associated operations on respective tables of the database to generate one or more results; and
send the one or more results to the stateful data processing service; and wherein

the stateful data processing service is further configured to, for the given client:
receive, from the stateless data processing service, the one or more results;
copy, from the database to a data cache, at least a portion of one or more other tables of the database to be used for one or more other table scans and associated operations of the plurality of table scans and associated operations;
perform the one or more other table scans and associated operations on the data cache to generate one or more other results;
generate a query result based at least on the one or more results and the one or more other results; and
send the query result to the client.

2. The system as recited in claim 1 , wherein to select, based on one or more criteria, the one or more table scans and associated operations from among the plurality of table scans and associated operations, the stateful data processing service is further configured to, for individual ones of the one or more table scans and associated operations:
determine that an amount of time to perform the table scan and associated operations by the stateless data processing service will be less than an amount of time to perform the table scan and associated operations by the stateful data processing service.

3. The system as recited in claim 1 , wherein to select, based on one or more criteria, the one or more table scans and associated operations from among the plurality of table scans and associated operations, the stateful data processing service is further configured to, for individual ones of the one or more table scans and associated operations, determine one or more of:
a size of a table to be scanned by the table scan is above a threshold size, or
a number of requests to be made by the stateless data processing service to perform the table scan is above a threshold number.

4. The system as recited in claim 1 , wherein to select, based on one or more criteria, the one or more table scans and associated operations from among the plurality of table scans and associated operations, the stateful data processing service is further configured to, for individual ones of the one or more table scans and associated operations, determine that:
no data of a table to be scanned by the table scan is stored by the stateful data processing service, or
an amount of data of the table to be scanned by the table scan that is stored by the stateful data processing service is less than a threshold amount.

5. The system as recited in claim 1 , wherein to select, based on one or more criteria, the one or more table scans and associated operations from among the plurality of table scans and associated operations, the stateful data processing service is further configured to, for individual ones of the one or more table scans and associated operations, determine that:
an amount of data returned by the table scan and associated operations is less than a threshold amount.

6. A method, comprising:
performing, by a plurality of computing devices of a provider network:
performing, by a stateful data processing service for a given client of the stateful data processing service:
receiving, from the client, a query for a database, wherein the query indicates at least one or more table scans and associated operations to be performed on one or more tables of the database, and wherein at least a portion of the one or more other tables is stored by a data cache of the stateful data processing service;
selecting, based on one or more criteria, at least one of the one or more table scans and associated operations;
sending an indication of the at least one table scan and associated operations to a stateless data processing service;

performing, by the stateless data processing service:
receiving, from the stateful data processing service, the indication of the at least one table scan and associated operations;
performing the at least one table scan and associated operations on respective tables of the database to generate one or more results; and
sending the one or more results to the stateful data processing service; and

performing, by the stateful data processing service:
receiving, from the stateless data processing service, the one or more results;
generating a query result based at least on the one or more results;
sending the query result to the client.

7. The method as recited in claim 6 , further comprising performing at least one other data scan on the data cache to generate one or more other results, and wherein generating the query result comprises:
generating the query result based at least on the one or more results and the one or more other results.

8. The method as recited in claim 6 , wherein selecting, based on one or more criteria, at least one of the one or more table scans and associated operations comprises:
determining a size of a table to be scanned by the at least one table scan is above a threshold size; and
determining a number of requests to be made by the stateless data processing service to perform the at least one table scan is above a threshold number.

9. The method as recited in claim 8 , wherein performing the at least one table scan and associated operations on respective tables of the database to generate one or more results comprises:
performing the number of requests in parallel to perform the at least one table scan.

10. The method as recited in claim 6 , wherein selecting, based on one or more criteria, at least one of the one or more table scans and associated operations comprises:
determining that an amount of data of the at least one table to be scanned by the table scan that is stored by the stateful data processing service is less than a threshold amount.

11. The method as recited in claim 6 , wherein selecting, based on one or more criteria, at least one of the one or more table scans and associated operations comprises:
determining that a table used by the at least one table scan is not used by other table scans of the query.

Assignees

• Amazon Tech Inc

Inventors

• Pandis Ippokratis
• Perron Matthew James

Classifications

• G06F16/24542

  Plan optimisation · CPC title

• G06F16/24537 Primary

  of operators · CPC title

• G06F16/24539

  using cached or materialised query results · CPC title