Databases, object storage, data platforms and durable application state.
Showing 1–10 of 10 dossiers
Supabase Pipelines turns Postgres WAL into a managed analytics feed for BigQuery. It isolates analytical workloads from production, but public-alpha pricing, Frankfurt-hosted pipeline infrastructure and destination constraints matter before adoption.
Supabase’s self-hosted stack now routes through Envoy by default, bringing new API-key support and hardened gateway defaults while breaking some Kong-specific assumptions.
R2’s new `us` jurisdiction gives object-storage users an explicit US data-residency guarantee, with jurisdiction-specific S3 endpoints and Workers bindings. Existing unrestricted buckets cannot simply be flipped into the new jurisdiction because jurisdiction is immutable after creation.
Aurora Serverless can now add roughly 12 ACUs in the first second of a scale-up event on platform versions 3 and 4. The change is automatic and is most useful for bursty SaaS, API, batch and agent workloads, but it does not remove the separate resume delay when a database has scaled all the way to zero.
Turso’s hosted early preview adds `BEGIN CONCURRENT` transactions backed by MVCC. Writes to different rows can proceed in parallel, while conflicting transactions fail at commit and must retry. The feature targets a core scaling constraint that often pushes applications away from SQLite-style architectures.
Neon’s beta backend now combines Postgres branches with Node.js Functions and S3-compatible Object Storage that inherit branch semantics. For builders, that makes ephemeral preview/test environments more complete: database state, backend code and object data can move together instead of requiring separate production-adjacent services.
The new tracing path gives Supabase requests a shared trace_id across browser or app spans and backend logs, with support for OpenTelemetry-compatible tooling. For JavaScript users, the practical detail is a required upgrade to supabase-js 2.112.0 or later because 2.106.0 through 2.111.x silently failed to propagate headers in bundled applications.
Retention-locked backups are gaining a project-level consequence: Google Cloud plans to create automatic liens that can block project deletion while protected backups remain. Infra teams need to account for this in teardown automation, IAM and recovery design.
Replit’s August 2026 Cloud pricing changes materially lower several production costs: autoscale compute falls from $3.20 to $0.60 per million compute units and database storage from $1.50 to $0.35 per GiB-month. The details matter because not every SKU moved down.
Cloud Storage project deletion no longer necessarily destroys every soft-deleted bucket immediately. Google’s August 17 change makes bucket retention part of project-recovery behavior, affecting disaster recovery, teardown assumptions and ongoing storage cost.
Data systems are difficult to replace precisely because they hold the part of an application that must survive. New databases and managed storage products can remove operational work, but compatibility, consistency, backup behaviour, egress and pricing matter more than a clean getting-started example.
This page follows consequential changes across relational databases, document stores, object storage and managed data platforms. BTN reads the technical details and looks at migration, failure recovery and the long-term cost of keeping data useful. Coverage is aimed at builders choosing a dependable default or evaluating a genuine new capability, not collecting database categories for their own sake.
The beat also covers replication, local-first systems and developer-friendly database services when they create a meaningful new option. Claims about effortless scale are checked against consistency, tooling and recovery. The safest data platform is often the one whose behaviour a team can explain during an ordinary failure.