NewApp Digest
search bot

processes

How and why we built Managed PostgreSQL for AI workloads
Tech Optimizer
October 10, 2026

How and why we built Managed PostgreSQL for AI workloads

The Nebius Managed PostgreSQL team developed a solution for AI workloads using CloudNativePG and Kubernetes, resulting in improved backup and restore times by transitioning from Barman to WAL-G, reducing a 1.5 TB backup duration from over a day to two hours. Since its launch in 2025, the service supports hundreds of production clusters and manages tens of tebibytes of data. Key features include high availability, point-in-time recovery, automated failover, compressed and encrypted backups, rapid restoration, connection pooling, and strict synchronous replication for zero data loss. Kubernetes and CloudNativePG were chosen for their advantages, including out-of-the-box features and open-source distribution. Each cloud project has a separate MK8s cluster for customer isolation and security. Persistent data is stored on fault-tolerant SSD drives, enabling swift instance migration during failures. Regular physical backups are archived, allowing restoration within a seven-day window with a single click. The transition from Barman to WAL-G addressed slow backup and restoration times and resource control limitations. With WAL-G, backups and restorations for a 1.5 TB cluster improved to two hours and one hour, respectively. PgBouncer was implemented to manage database connections effectively, reducing connection storms during peak AI workload traffic. Two vector search extensions, pgvector and pgvectorscale, are available to cater to different memory and dataset needs. Benchmarking showed that Nebius performs competitively with AWS at a lower cost. The infrastructure integrates cloud engineering principles with specialized expertise to meet AI workload demands.
DBtune autonomous optimization blows PostgreSQL bottlenecks away
Tech Optimizer
October 10, 2026

DBtune autonomous optimization blows PostgreSQL bottlenecks away

DBtune, a startup based in Malmö/Lund, Sweden, was founded in 2020 by Dr. Luigi Nardi as a spin-off from Stanford University and Lund University. The company focuses on enhancing PostgreSQL's capabilities through autonomous database optimization technology and has raised .4 million in a seed round from various investors in July 2023, with support from the Wallenberg Foundation and Vinnova. DBtune's technology utilizes agentic AI for optimizing PostgreSQL databases by tuning server parameters, optimizing indexes, and monitoring autovacuum processes. The release of DBtune version 4.0 incorporates machine learning and AI to dynamically adjust server parameters and suggest optimal indexing strategies. The platform identifies and ranks potential index opportunities based on workload and provides recommendations for new indexes and maintenance of existing ones. A case study with Midwest Tape demonstrated a 10.8x performance improvement in database query runtime after employing DBtune for performance tuning. Benchmarking tests showed a transaction throughput increase of 87% during initial optimization, with a further 17% improvement on an already optimized system. Disk I/O activity was reduced by 43%, and WAL generation decreased by 23%. Dr. Nardi emphasizes the need for AI-driven optimization to manage the complexities of PostgreSQL databases and rising cloud costs, suggesting that significant code adaptations will be necessary for achieving fully autonomous PostgreSQL.
How Telangana’s MeeSeva migration to open source could make governments rethink software costs
Tech Optimizer
October 10, 2026

How Telangana’s MeeSeva migration to open source could make governments rethink software costs

In April 2026, Telangana's MeeSeva platform migrated from the Oracle database to the open-source PostgreSQL database after 15 years of reliance on Oracle. This transition aimed to reduce annual costs of approximately ₹10 crore associated with Oracle licensing and maintenance. The initiative began in 2023, and the migration was executed with less than four hours of downtime, processing around 80,000 transactions daily. The migration involved over 85 microservices, three Oracle servers, and 1,939 tables, with the process automated by the Ora2Pg tool, which handled 60-70% of the transition. The migration cost was below ₹10 lakh, and after the switch, MeeSeva processed nearly 200 million transactions without data loss. The new architecture includes automatic failover capabilities to minimize disruption.
Microsoft admits one bad driver can crash all of Windows, and Intel’s Windows 11 26H2-ready GPU drivers are next to get its fix
Winsage
October 9, 2026

Microsoft admits one bad driver can crash all of Windows, and Intel’s Windows 11 26H2-ready GPU drivers are next to get its fix

Microsoft is collaborating with Intel, AMD, and Nvidia to improve driver quality for Windows 11, addressing issues that can lead to system failures. Intel has confirmed that its GPU drivers are compatible with Windows 11 26H2, and recent updates have integrated support for this version. The Driver Quality Initiative (DQI) aims to enhance driver stability and quality across the ecosystem. Microsoft emphasizes that high-quality drivers are essential for reliable Windows performance and that a collective effort from all stakeholders is necessary. The DQI focuses on four principles: Architecture, Trust, Lifecycle, and Quality Measures, promoting user-mode drivers, stricter certification, removal of outdated drivers, and assessments based on stability, functionality, performance, power consumption, battery life, and thermal performance.
Windows 11 26H2 Update Lowers RAM Consumption
Winsage
October 6, 2026

Windows 11 26H2 Update Lowers RAM Consumption

Microsoft's Windows 11 26H2 update has introduced optimizations that reduce RAM consumption during idle periods and improve RAM management during active use. Users have reported a decrease in idle RAM usage by approximately 2 to 3 GB, with testing showing a drop from 10 GB to 7 GB on a laptop with 16 GB of RAM. The update also includes the Low Latency Profile feature, which boosts CPU performance for 1 to 3 seconds, resulting in faster application launch times—up to 40% faster for Microsoft applications like Edge and Outlook, and up to 70% for elements like the Start Menu and context menus.
How to supercharge your Windows 11 PC with this built-in automation tool
Winsage
October 6, 2026

How to supercharge your Windows 11 PC with this built-in automation tool

Windows 11 includes a Task Scheduler that automates repetitive tasks to enhance productivity. To set up a task, users should open Task Scheduler, select "Create Task," and fill in details under General, Triggers, Actions, Conditions, and Settings. It's important for laptop users to avoid scheduling tasks that run every five minutes to maintain low-power idle states. The "Run with highest privileges" option should be used cautiously for security reasons. Three recommended automations include: 1. Using WinGet to keep software updated with the command winget upgrade --all --silent --accept-package-agreements --accept-source-agreements --disable-interactivity, scheduled weekly. 2. Disabling unnecessary startup applications and launching them with a delayed logon trigger to reduce startup clutter. 3. Automating tasks with PowerShell for file transfers or system maintenance, ensuring they run during idle times or when connected to AC power. Windows also performs some maintenance tasks automatically, such as deleting temporary files and optimizing drives.
Devart rolls out dbForge 2026.2
Tech Optimizer
October 6, 2026

Devart rolls out dbForge 2026.2

Devart has released dbForge 2026.2, a database management suite update featuring a PostgreSQL Debugger for error tracking, visual editors for managing PostgreSQL objects, and an enhanced dbForge AI Assistant with broader context awareness. The AI Assistant now supports models from Anthropic and OpenAI and includes a free Express edition with limited features. The update also improves SQL development and schema comparison functionalities, with new refactoring commands and support for MySQL 26.7 and MariaDB 13.x. dbForge 2026.2 is available for download, with current users able to update through the Help menu.
RuneScape dev apologises after RuneFest debut of an AI-generated six-fingered character: 'That's pretty hard to explain away'
AppWizard
October 6, 2026

RuneScape dev apologises after RuneFest debut of an AI-generated six-fingered character: ‘That’s pretty hard to explain away’

Jagex hosted its annual RuneFest in Birmingham, UK, where it unveiled a trailer for the Dragon Slayer 2 quest. The trailer featured noticeable flaws, such as a character with six fingers and unusual sneaker-like soles on their armoured shoes. Social media users pointed out the apparent use of generative AI in the trailer, prompting backlash against Jagex, especially after a recent controversy involving a clothing collaboration with 100 Thieves. Alec 'Mod Sween' Sween from Old School RuneScape acknowledged the issues, stating that the trailer had been taken down for urgent investigation. He clarified that it was produced by a trusted external partner who assured Jagex that generative AI would not be used. Sween emphasized the company's commitment to quality and announced plans to overhaul their review procedures and collaborate with community creators to ensure high standards for Dragon Slayer 2. He admitted that the team fell short of expectations and expressed concern over the incident highlighting weaknesses in their internal processes.
Search