@this-npm-test-org/connection-aws-cloudwatch
v0.1.3
Published
Monitor AWS metrics, alarms, logs, and dashboards.
Readme
AWS CloudWatch
Monitor AWS metrics, alarms, logs, and dashboards.
Install
amodal connect aws-cloudwatchAuthentication
AWS Signature Version 4 signing process.
AWS_ACCESS_KEY_ID— IAM access keyAWS_SECRET_ACCESS_KEY— IAM secret keyAWS_REGION— AWS region (e.g., us-east-1)AWS_SESSION_TOKEN— (optional) for temporary credentials (STS, IAM roles)
IAM policy must grant cloudwatch:GetMetricData, cloudwatch:DescribeAlarms, logs:GetLogEvents, etc.
Endpoints
Metrics
POST / Action=GetMetricData— Get metric data (batch, up to 500 metrics)POST / Action=GetMetricStatistics— Get statistics for a metricPOST / Action=ListMetrics— List available metrics
Alarms
POST / Action=DescribeAlarms— List alarms and their statesPOST / Action=DescribeAlarmHistory— Get alarm state change historyPOST / Action=PutMetricAlarm— Create or update alarm (confirmation)POST / Action=DeleteAlarms— Delete alarms (confirmation)POST / Action=SetAlarmState— Manually set alarm state (confirmation)
Dashboards
POST / Action=GetDashboard— Get dashboard definitionPOST / Action=ListDashboards— List dashboardsPOST / Action=PutDashboard— Create or update dashboard (confirmation)
Logs
POST /logs Action=GetLogEvents— Get log events from a streamPOST /logs Action=FilterLogEvents— Search log events with patternPOST /logs Action=DescribeLogGroups— List log groups
Entities
- log
Rules
- Rate limits vary by action: GetMetricData (50 TPS), GetMetricStatistics (400 TPS), DescribeAlarms (9 TPS), FilterLogEvents (5 TPS per log group)
- Metric data retention depends on resolution: 1-second (3 hours), 1-minute (15 days), 5-minute (63 days), 1-hour (455 days)
- CloudWatch Metrics and Logs are separate services with different endpoints — logs use
logs.${AWS_REGION}.amazonaws.com - Cross-region queries are not supported — must query each region separately
