Table Dashboard
Table metadata, key schema and 24 hours of CloudWatch metrics (consumed capacity, latency, throttles) plus active alarms in a slide-in panel.
Opening the dashboard
Click Dashboard in the table info bar. A panel slides in with a 🔄 refresh button and a ✕ to close.
Table metadata
The top grid comes straight from DescribeTable and updates on every refresh.
| Stat | Meaning |
|---|---|
| Status | ACTIVE in green, CREATING or UPDATING in amber, anything else in red. |
| Item Count | Estimated by AWS and refreshed roughly every six hours. |
| Table Size | Total bytes, shown in B, KB, MB or GB. |
| Billing | On-Demand (PAY_PER_REQUEST) or Provisioned. Provisioned tables also show Prov. RCU and Prov. WCU. |
| GSIs / LSIs | Number of global and local secondary indexes. |
| Streams | ON or OFF, mirroring the table’s stream specification. |
| Replicas | Number of global-table replicas. |
| Created | Table creation date. |
Below the grid, the Key Schema section lists each key attribute with its type (String, Number or Binary) and role (Partition Key or Sort Key).
CloudWatch — last 24 hours
Five sparkline cards summarise the previous 24 hours at one-hour resolution. Each card shows the 24-hour total and an hourly average; a series with no datapoints reads “No data”.
- Consumed RCU and Consumed WCU — sums per hour.
- Latency (ms) — average SuccessfulRequestLatency.
- Read Throttles and Write Throttles — sums per hour.
CloudWatch is queried with the connection’s region and credentials but never through a custom endpoint. Against DynamoDB Local or LocalStack, or when the IAM principal lacks CloudWatch permissions, the section explains that metrics are unavailable for this endpoint.
dynamodb:DescribeTable. Metrics and alarms need cloudwatch:GetMetricStatistics and cloudwatch:DescribeAlarmsForMetric.{
"Effect": "Allow",
"Action": [
"dynamodb:DescribeTable",
"cloudwatch:GetMetricStatistics",
"cloudwatch:DescribeAlarmsForMetric"
],
"Resource": "*"
}Active alarms
When CloudWatch alarms are attached to the table’s ConsumedReadCapacityUnits metric, an Active Alarms section lists them with their state: OK, ALARM or INSUFFICIENT_DATA, colour-coded. Alarms on other metrics are not listed. The section is hidden when there are no alarms.