Dasha 1.10 is an open‑source performance dashboard for PostgreSQL fleets that connects to clusters using a read‑only role, displaying real‑time activity without installing agents or extensions on database hosts. One instance can serve multiple clusters and their replicas, aggregating per‑instance statistics from each host. A new feature in this release is Patroni auto‑discovery: Dasha reads Patroni keys stored in its DCS (etcd v3 or Consul) and treats each Patroni scope as a cluster, mapping members to hosts and databases to PostgreSQL instances. Changes such as new clusters, lost replicas, and database creation/deletion appear within the refresh interval. DCS interaction is read‑only; TLS with client certificates is supported for both etcd and Consul. The UI shows standby relationships on Summary and Replication pages, matching clusters by PostgreSQL system identifier even when a standby is behind a VIP. Activity & Waits (PostgreSQL 14+) visualizes live session charts by class (CPU, IO, Lock, LWLock, etc.) and idle transactions, sampling pg_stat_activity every 1‑10 seconds. It also lists top statements, sessions, databases, users, applications, and wait events, with drill‑down filters and query reports. Replication slot invalidation forecasts measure WAL rate over configurable windows (30 seconds to 5 minutes) and display retained WAL, remaining max_slot_wal_keep_size reserve, and estimated time until slot loss, using two LSN samples without maintaining connections. COPY progress for PostgreSQL 14+ shows tables, source type, bytes/rows processed, exclusions, and ETA. Health Score now works with plain exporters (postgres_exporter, node_exporter, pgbouncer_exporter) and automatically maps Patroni cluster metrics by scope and member name. Cold tables are excluded from score penalties, and a separate recommendation lists tables needing a single VACUUM (FREEZE, ANALYZE). Additional features include query analysis snapshots, index advisor with B‑tree suggestions, and bloat/duplication detection. Business relevance is that system integrators and ERP operators can leverage Dasha to monitor large PostgreSQL environments without agent deployment, gain immediate visibility into Patroni‑managed clusters, predict replication slot exhaustion to avoid downtime, and receive actionable performance recommendations—all integrated into existing Prometheus/VictoriaMetrics stacks, supporting proactive maintenance and reducing operational overhead.
Relevansi untuk bisnis Anda
For organizations using Pramadya’s system integration services, Dasha 1.10 provides a no‑agent, PostgreSQL‑agnostic monitoring solution that automatically discovers Patroni‑orchestrated clusters, aligns with existing cloud‑native infrastructure, and delivers predictive alerts for replication slot loss. This capability reduces the risk of unexpected downtime, streamlines health scoring across multiple instances, and offers actionable index and vacuum recommendations. By integrating with standard exporters, Dasha fits seamlessly into current observability pipelines, enabling Pramadya’s customers to maintain high‑performance ERP systems with minimal administrative overhead and faster troubleshooting cycles.
Baca sumber asli: PostgreSQL News
Diskusikan kebutuhan sistem Anda bersama Pramadya.
Sumber: PostgreSQL News