Monitor Elastic IP Address (EIP) metrics such as bandwidth and packet rates, set up threshold-based alert rules, and subscribe to abnormal event notifications to track EIP usage and respond quickly to issues.
Monitoring tools at a glance
EIP provides three complementary monitoring tools. Choose based on your use case:
| Tool | Best for |
|---|---|
| CloudMonitor | Standard minute-level metrics, threshold alerts, and monitoring charts |
| Fine-grained monitoring | Second-level metrics to catch traffic spikes before throttling occurs |
| Abnormal events | Reviewing and subscribing to network anomaly events from the past 30 days |
CloudMonitor alerting
CloudMonitor (CMS) collects EIP metrics in real time. Use it to:
Check EIP health on monitoring charts.
Create threshold alert rules to detect anomalies and take action.
RAM user access: If a Resource Access Management (RAM) user needs to view EIP monitoring data, grant the required permissions using an Alibaba Cloud account first:
Option 1: Attach a system policy that includes the
cms:QueryMetricListaction to the RAM user. Available system policies: AliyunCloudMonitorFullAccess, AliyunCloudMonitorMetricDataReadOnlyAccess, AliyunCloudMonitorReadOnlyAccess, or AliyunCloudMonitorReadOnlyBasicAccess.Option 2: Create or modify a custom policy with
cms:QueryMetricListin an Allow statement, then attach the policy to the RAM user.
EIP metrics
| Metric | Description |
|---|---|
| Outbound Bandwidth | Outbound traffic bandwidth (bps) |
| Inbound Bandwidth | Inbound traffic bandwidth (bps) |
| Outbound Bandwidth Usage | Outbound bandwidth as % of capacity |
| Inbound Bandwidth Usage | Inbound bandwidth as % of capacity |
| Outbound Packet Rate | Packets sent per second (pps) |
| Inbound Packet Rate | Packets received per second (pps) |
| Rate of Outbound Packets Dropped | Outbound packets dropped per second due to throttling (pps) |
| Rate of Inbound Packets Dropped | Inbound packets dropped per second due to throttling (pps) |
View metrics
EIP console
Go to the Elastic IP Addresses console, click the target EIP ID, then click the Monitoring and O&M > Monitor tab.
Default time range: last hour. Switch to the last 3, 6, or 12 hours, or specify a custom range. Data is retained for up to 31 days.
Turn on Auto Refresh for minute-by-minute updates.
CloudMonitor console
Go to the CloudMonitor console - Elastic IP Address page, find the target EIP, and click Monitoring Charts in the Actions column.
Default time range: last hour. Switch to the last 3 hours, 6 hours, 12 hours, 1 day, 3 days, 7 days, or 14 days, or specify a custom range. Data is retained for up to 31 days.
API
Call DescribeMetricList to query monitoring data for a specific metric.
Set up alerts
Releasing an EIP also deletes its alert rules in CloudMonitor.
Console
Go to the CloudMonitor console - Alert Rules page and click Create Alert Rule. Configure the following settings:
Product: Select Elastic IP Address.
Resource Range: Choose All Resources to apply the rule to every EIP under the account, or choose Instances and configure Associated Resources to target specific EIPs.
The resource range cannot be changed after the rule is created.
Rule Description: Click Add Rule, select a metric type, then configure the Alert Rule, Metric Type, and Metric in the Configure Rule Description panel.
Mute Period: How often to resend the notification if the issue persists.
Effective Period: The time window during which the rule actively checks metrics.
Alert Contact Group: The contact group to notify.
API
Call PutResourceMetricRules to create threshold alert rules for specific metrics of a resource.
Fine-grained monitoring
Standard monitoring uses minute-level granularity. Fine-grained monitoring increases this to per-second intervals, giving you visibility into traffic spikes that minute-level data averages out.
When bandwidth utilization reaches 100%, the EIP throttles traffic and drops packets, degrading client-side experience. Per-second monitoring lets you detect spikes and adjust bandwidth before throttling occurs.
Data is stored in Simple Log Service (SLS) for dashboards and alerting.
Rate-limiting drop data is only available for EIPs associated with NAT gateways or SLB instances. EIPs on ECS instances or ENIs don't report this metric because rate-limiting happens at the ECS NIC level.
Fine-grained metrics
| Metric | Description |
|---|---|
| Second-level bandwidth peak (outbound/inbound) | Per-second peak bandwidth in the outbound and inbound directions. Unit: Bps. |
| Second-level packet rate (outbound/inbound) | Packets sent and received per second in each direction. Unit: pps. |
| Second-level drop rate (outbound/inbound) | Packets dropped per second in each direction. Unit: pps. |
| TCP new connection rate (outbound/inbound) | New TCP connections established per second in each direction. Unit: pps. |
Enable fine-grained monitoring
Console
Go to the VPC console - Internet Tool Kit page and click Fine-grained Monitoring.
If the required permissions are not granted, follow the on-screen wizard to authorize cloud resource access. RAM users need the Alibaba Cloud account to complete authorization first.
Find the target EIP and click Enable Fine-grained Monitoring in the Actions column. Select an existing log project and Logstore, or create new ones, to complete the log configuration.
After the feature is enabled, click the
icon in the Monitor column to view and analyze monitoring data.
API
Call SetHighDefinitionMonitorLogStatus to enable or disable fine-grained monitoring for an EIP.
Billing
Enabling fine-grained monitoring sets the Logstore retention to 7 days. Each account gets 500 MB of free storage per month — usage beyond this incurs charges.
Abnormal events
EIP integrates with the Event Center of Network Intelligence Service (NIS) to surface network anomalies and their recommended fixes.
After an EIP is associated with a cloud resource, go to the EIP details page and click the Monitoring and O&M > Event Center tab. All abnormal events from the last 30 days are available for review.
From the Event Center, you can:
Follow the recommended actions to resolve issues.
Subscribe to events to receive alert notifications for critical events.