data retrieval

Tech Optimizer
August 5, 2026
Google Cloud has introduced its Database Migration Service, which focuses on converting SQL Server stored procedures with multiple result sets into PostgreSQL code. The service uses an automated decision-making process to determine if a SQL Server procedure should be translated into a PostgreSQL stored procedure or function, based on the number of result sets and the presence of a scalar return value. Procedures with a single result set or scalar return value are converted into PostgreSQL stored procedures, while those with multiple result sets are transformed into functions returning a SETOF refcursor. In a healthcare reporting example, a master procedure retrieves patient details and may return multiple result sets along with a status integer. The PostgreSQL translation involves simpler child procedures as standard procedures and more complex routines as functions that open cursors sequentially. Scalar return values are handled by placing them in a dedicated cursor at the end of execution, changing how applications interact with the outputs. Testing of these translated objects must occur within an explicit transaction block due to PostgreSQL cursor lifecycle constraints. The user executes the function to generate cursor references and fetches data sequentially. Google Cloud's internal analysis classifies stored procedures by scanning for direct result sets and considers conditional logic and loops that may complicate result set counts. A directed graph of procedure calls is constructed to determine the result set classification. Database migrations often face challenges due to legacy application logic, especially with stored procedures designed to minimize database round trips. Successful code conversion is only part of the task, as teams must also adjust test harnesses and application data access layers for PostgreSQL's cursor management. Google Cloud's service categorizes SQL Server procedures into three types: no result sets, a single result set, or multiple/dynamic result sets.
Tech Optimizer
July 29, 2026
EDB Postgres AI is a pioneering solution that integrates intelligence and data on a unified sovereign foundation, eliminating the need for ETL processes, data duplication, and separate vector stores. Independent benchmarks show that EDB Postgres AI outperforms competing platforms in speed, accuracy, and cost-effectiveness. A study by McKnight Consulting Group indicates that EDB Postgres AI excels in key performance metrics, including query latency, accuracy, cost, and data freshness, outperforming specialized vector databases and other managed Postgres platforms. EDB Postgres AI achieves median query latencies of 50 milliseconds at a scale of 50 million vectors, making it 80x faster than Databricks, 21x faster than MongoDB Atlas, and up to 2x faster than alternatives like Aurora and Crunchy Bridge. It also delivers the highest recall rates for core vector searches, surpassing competitors like MongoDB and Databricks. In tests involving concurrent retrievals, EDB Postgres AI completed a three-arm agent loop in 27 milliseconds, significantly faster than other platforms. In terms of cost efficiency, EDB Postgres AI offers 76x better price performance than Databricks, 34x better than MongoDB, and 23%–28% better than the nearest managed Postgres competitors when considering query speed. The architecture of EDB Postgres AI allows for the integration of vector, filtered, and full-text retrieval in a single query path, ensuring real-time data access and eliminating the need for separate systems.
Tech Optimizer
June 24, 2026
EnterpriseDB is addressing challenges in AI development projects, particularly data sprawl, by introducing features in the EDB Postgres AI platform. The platform now includes Converged Analytics, which bridges operational and analytical data without complex ELT pipelines, and the Agentic Database, which transforms the system into an autonomous database that proactively manages over 200 metrics. These innovations aim to consolidate various data types into a single governed platform, reducing complexity and costs associated with database administration. The update also introduces governance capabilities at the data layer, expected to be available in the latter half of 2026, and a bring-your-own-cloud option for applying AI to data. Customer feedback has influenced these developments, highlighting the need for reduced manual intervention in data management.
Tech Optimizer
June 20, 2026
PostgreSQL version 18 has deprecated MD5 password authentication in favor of SCRAM-SHA-256, with a new parameter, md5_password_warnings, enabled by default to log deprecation warnings. It has enhanced monitoring capabilities by adding columns to pg_stat_database and pg_stat_statements to track parallel worker activity, with the default max_parallel_workers_per_gather set to 0 in Aurora PostgreSQL. The pg_stat_subscription_stats view now includes new columns for tracking conflict types in logical replication. Optimizer statistics are automatically transferred during upgrades, while uuidv7() generates timestamp-ordered UUIDs. The default streaming option for CREATE SUBSCRIPTION has changed to parallel, and the idle_replication_slot_timeout parameter automatically invalidates inactive replication slots. Enhancements to the COPY command include REJECT_LIMIT for error tolerance and a silent LOG_VERBOSITY level. OLD and NEW aliases have been introduced in RETURNING clauses for various DML commands.
Search