ClickHouse Postgres: Unified Data, Smarter Ops

Bridging OLTP and OLAP: A Unified Data Vision

The latest updates to ClickHouse Managed Postgres demonstrate a strong commitment to realizing its vision of a unified data stack for both OLTP and OLAP workloads. The emphasis on a simplified onboarding experience, coupled with enhanced observability through logs, metrics, and a Prometheus endpoint, directly addresses common pain points for database administrators and developers. The proactive customer notifications for disk usage, with upcoming additions for replica lag and uptime, are a significant step towards operational excellence, allowing users to stay ahead of potential issues. Furthermore, the expansion of extension support, particularly with pg_re2 offering performance parity with ClickHouse's regex capabilities within Postgres, and improvements to pg_clickhouse for seamless cross-database querying, are crucial for developers seeking to leverage the strengths of both systems without compromising performance or consistency.

However, while the focus on performance improvements for backups and the introduction of extensions like pg_re2 are commendable, the article could benefit from more quantitative data on the performance gains achieved. For instance, providing benchmarks for the faster backups or specific latency improvements for queries utilizing pg_clickhouse would offer a clearer picture of the impact. The article mentions the upcoming migration to WAL-RUS for backups, which is a positive step towards resource efficiency, but details on the expected benefits of this transition would be valuable. The unified querying layer via pg_clickhouse is a core differentiator, but the article could elaborate further on the complexity of setting up and managing these cross-database joins, especially in large-scale environments. The roadmap hints at deeper OpenTelemetry support, which is a welcome addition for integrating into existing observability stacks, but concrete timelines or examples would enhance its perceived value.

Key Points

• ClickHouse Managed Postgres has enhanced its onboarding experience for quicker setup and value realization.
• Proactive customer notifications are now available for disk usage (above 85%), with upcoming alerts for replica lag and uptime.
• Observability has been significantly improved with direct console access to PostgreSQL logs, refreshed metrics (including per-table breakdowns and query insights), and a Prometheus-compatible /metrics endpoint.
• Backup and restore speeds have been increased by leveraging more CPU and disk throughput on larger server sizes.
• Extension support has expanded with pg_re2 for faster regular expressions in Postgres and continued improvements to pg_clickhouse for unified querying and pg_stat_ch for query pattern monitoring.
• Future plans include GCP availability, managed maintenance, broader configuration controls, and enhanced data integration with cloud object stores.

---

📖 Source: What's new in ClickHouse Managed Postgres: Customer notifications, better observability, faster backups, extensions, and more