Grafana PostgreSQL RDS CloudWatch Dashboard
验证：未记录
- 适用版本
- 未记录
- 验证范围
- 尚未登记验证记录。
本文缩写（英文全称与中文含义）
| 缩写 |
英文全称／名称 |
中文含义 |
| DB |
Database |
数据库 |
| I/O |
Input/Output |
输入／输出 |
| IOPS |
Input/Output Operations Per Second |
每秒输入输出操作数 |
| RDS |
Relational Database Service |
AWS 关系型数据库服务 |
| SQL |
Structured Query Language |
结构化查询语言 |
Links#
1. Dashboard Rows#
| Row |
CloudWatch Metrics |
Statistic |
Display |
| Overview |
CPUUtilization |
Average |
time series by DB instance |
| Overview |
FreeableMemory |
Average |
memory headroom |
| Overview |
DatabaseConnections |
Average |
connection pressure |
| Overview |
ReadLatency, WriteLatency |
p95 / Average |
storage latency |
| Drilldown |
ReadIOPS, WriteIOPS |
Average |
I/O pressure |
| Drilldown |
Deadlocks |
Sum |
DB contention signal |
| Drilldown |
AuroraReplicaLagMaximum |
Maximum |
Aurora replica health |
| Capacity |
VolumeBytesUsed |
Average |
Aurora storage used trend |
| Debug |
latency split and IOPS split |
Average / p95 |
storage troubleshooting |
2. Notes#
RDS CloudWatch 反映平台资源；SQL、锁和慢查询定位仍需 postgres_exporter / pg_stat_statements 等数据库内信号。
CloudWatch dashboard 只能覆盖 RDS 平台层。
PostgreSQL SQL 级别问题仍然需要 postgres_exporter / pg_stat_statements。
如果你当前看的对象是 Aurora PostgreSQL cluster，优先使用:
VolumeBytesUsed
AuroraReplicaLagMaximum
如果 cluster 没有 reader，replica lag panel 可能没有数据。