Observable Graphics: Clarity and Vision

Heaven's movement, all-seeing view, perceiving details to master the whole!

Infra Built-in: Built-in monitoring infrastructure

Pigsty provides a ready-to-use Prometheus & Grafana observability tech stack infrastructure

Works out of the box with automatic monitoring discovery. Enterprise-grade monitoring capabilities without additional configuration

Time-series database: Prometheus / AlertManager / X Exporters
Visualization dashboards: Grafana, Echarts, plus ten additional panels/data sources
Log collection: Loki / Promtail, or VictoriaMetrics / Vector

Monitoring systems built with data analytics and BI principles, transforming thousands of metrics into dozens of carefully designed dashboards

From macro overview to micro details, system status at a glance, providing solid data support for management

3000+ metrics, 666 pre-aggregation rules, collecting all possible observation points
Unified label system for cross-component correlation, multi-level visualization panels with interactive linkage
Laying solid data foundation for intelligent operations — 《D-Bot LLM as DBA》

Five years of refinement for the ultimate PG monitoring experience. Global/cluster/instance/DB/object five-level linkage

Detailed down to historical metrics of individual tables, indexes, functions, and query objects in a database instance

Global/cluster/instance/database/object five-level linkage, one-click drill-down, efficient navigation
Support for reviewing historical detailed metrics of tables, indexes, functions, queries and other database objects
Self-developed pg_exporter, compatible with pgbouncer, customizable metrics for in-database APM

Can unify management of hosts, databases, applications, and LB monitoring with correlation analysis, building a full-stack operations perspective

Supports monitoring cloud vendors' RDS and PG-compatible services like PolarDB / Aurora
Built-in monitoring for all modules: Node, Infra, MinIO, Etcd, Redis, Haproxy, Nginx…
Can serve as a general monitoring solution for servers and other applications, easy to integrate

Production-tested preset alert rule sets, zero configuration, automatic critical event notification

Pre-aggregated calculation of key metrics, fault-specific dashboard presentations, improving problem localization efficiency

AlertManager integration: aggregation, silencing, IM connectivity
56 preset alert rules covering all built-in Pigsty modules
Alert information links directly to monitoring dashboards, improving fault diagnosis efficiency

Combining monitoring metrics with system catalogs to quickly identify and optimize resource-intensive slow queries

Intuitively check and evaluate query optimization effects from monitoring dashboards, generating quantifiable comparison results

pg_stat_statements collects and presents historical records of slow query key metrics
auto_explain automatically records slow query execution plans and visualizes with PEV
Use PGCAT application to directly read and present PG system view related data

Unified log collection system built with Loki and Promtail, centrally storing system and database logs

Search and filter all logs as easily as using a search engine, significantly improving efficiency in troubleshooting and root cause analysis

All log streams gathered in one place and associated through labels, flexible combination of search criteria
Use PromQL to extract additional monitoring metrics from logs, such as error rates
Use PGLOG application for interactive analysis of specific PG CSV log samples