When bandwidth usage spikes or packets are dropped, catching the issue early prevents service degradation. CloudMonitor (CMS) collects real-time metrics for your Internet Shared Bandwidth instance and plots them as time-series charts. You can also configure threshold-triggered alerts so that CloudMonitor notifies you automatically when a metric crosses your defined threshold.
View monitoring data
CloudMonitor samples metrics at regular intervals and retains historical data for up to one month. You can view data for the last 1 hour, 3 hours, 6 hours, 12 hours, or a custom time range within that period.
There are two ways to access the monitoring charts.
From the Internet Shared Bandwidth console
Log on to the Internet Shared Bandwidth console. In the top menu bar, select the region where your instance is located.
Open the monitoring charts for the target instance using either of the following methods:
Click the
icon in the Monitor column.Click the instance ID. On the Basic Information tab, scroll to the Monitor area.
From the CloudMonitor console
Log on to the CloudMonitor console. In the left navigation pane, click Cloud Service Monitoring.
Click Internet Shared Bandwidth. Select the region where the instance is located, then click Monitoring Charts in the Actions column of the target instance.
Monitoring metrics
Internet Shared Bandwidth exposes eight metrics in CloudMonitor. The bandwidth usage metrics (percentage values) are the most actionable for capacity planning and alert configuration.
Metric | Description |
Inbound Bandwidth | Bandwidth consumed by inbound traffic from external sources to the instance |
Outbound Bandwidth | Bandwidth consumed by outbound traffic from the instance to external destinations |
Inbound Packet Rate | Number of packets received by the instance per second |
Outbound Packet Rate | Number of packets sent by the instance per second |
Outbound Bandwidth Utilization | Outbound bandwidth consumption as a percentage of the total instance bandwidth |
Inbound Bandwidth Utilization | Inbound bandwidth consumption as a percentage of the total instance bandwidth |
Outbound Rate-Limited Packet Drop Rate | Rate at which outbound packets are dropped due to outbound rate limiting |
Inbound Rate-Limited Packet Drop Rate | Rate at which inbound packets are dropped due to inbound rate limiting |
FAQ
Why does packet loss occur even when bandwidth utilization has not reached the instance specification limit?
CloudMonitor displays bandwidth metrics as minute-level averages. Internet Shared Bandwidth enforces rate limiting based on per-second traffic.
If a traffic burst exceeds the instance bandwidth specification for just a few seconds, the rate limiter drops packets during that window—even if the 1-minute average stays below the limit. This is the most common reason for unexpected packet loss when monitoring charts show usage within normal range.
Recommendations:
Monitor second-level metrics to detect short-duration traffic bursts that minute-level averages can mask.
Reduce peak traffic or upgrade the instance specification to prevent bursts from triggering rate limiting.
Why does cross-border egress address traffic exceed the cross-border bandwidth specification without triggering errors?
EIP inbound bandwidth is driven by your workloads. Download-intensive applications, for example, can receive more traffic than your cross-border bandwidth specification allows. Traffic that exceeds the cross-border bandwidth limit is silently dropped—no errors are raised—so CloudMonitor metrics can show EIP inbound traffic higher than the cross-border specification simultaneously with the cross-border link operating at its limit. Both readings are accurate: the EIP receives the traffic, and the cross-border link discards whatever exceeds its cap.
Recommendations:
Set your Internet Shared Bandwidth peak to match your cross-border bandwidth specification to avoid silent traffic drops and unexpected charges.
What should I do if the Describe95Traffic API returns all zeros for traffic data?
Describe95Traffic only supports querying traffic data from the last two days (48 hours). If the query time range is earlier than two days before the current time, the API returns all zeros.
Solution: Adjust the query time range to within the last 48 hours to retrieve valid data.
Create a threshold-triggered alert rule
A threshold-triggered alert rule monitors your instance continuously. When a metric crosses the threshold you define, CloudMonitor sends an alert notification—so you can act on anomalies without watching the charts.
If you release an Internet Shared Bandwidth instance, CloudMonitor deletes all threshold-triggered alert rules associated with that instance.
Steps:
Log on to the CloudMonitor console.
In the left navigation pane, choose . Click Create Alert Rule and set Product to Internet Shared Bandwidth. For configuration details, see Create an alert rule.