Traffic probe

How it works

Traffic Probe is a non-intrusive network diagnostics feature that lets you test connectivity between your VPC and data centers. This service requires no agent installation or software deployment on your servers, providing a lightweight method to validate network paths.

1. Initialize the probe source
   When you create a Traffic Probe task, the system automatically creates a managed Elastic Network Interface (ENI) within your VPC and vSwitch. Ensure your selected vSwitch has sufficient available IPs before configuring a task. This ENI:

   • Serves as the probe executor.

   • Consumes one IP address from your vSwitch's available IPs.

   • Uses this allocated IP as the source for all probe packets.

2. Send and receive probe packets
   The ENI sends ICMP echo request packets, known as ping packets, to the destination IP at the specified probe frequency, such as every 15 seconds.

3. Determine connectivity
   Traffic Probe determines connectivity status based on response timing:

   • Reachable: ICMP echo reply received within the configured timeout period.

   • Unreachable: No ICMP echo reply received within the timeout period, indicating potential network blockage, routing issues, or destination unavailability.

4. Configure security policies
   To ensure the probe runs, complete the following configurations:

   • Associate a security group with the probe ENI when specifying your VPC and vSwitch.

   • Allow ICMP traffic:

     • Source: The outbound rules of the associated security group must allow ICMP traffic to the destination IP.

     • Destination: Security group rules at the destination allow inbound ICMP traffic from the probe source IP. This may include security groups, network ACLs, OS firewalls, or on-premises IDC firewalls.

Create a probe task

Before you begin, ensure that your VPC and IDC are connected and that you have activated network analysis and monitoring.

1. Go to the Traffic Probe in the Express Connect console and click Create Task.

2. On the Create Task > Service Availability Test page, configure the following parameters:

   • Basic Information:

     • Address: Enter the probe destination. This can either be an IP address, such as an IP of a server in your IDC, or a domain name that can be resolved properly.

     • Number of Ping Packets: Ping packets sent in parallel for each probe. The default is 10.

   • Select Detection Points:

     • Region: Where the probe source is located.

     • VPC: The VPC that is used as the probe source. A managed ENI will be automatically created in this VPC and it will send ping requests to the destination.

     • vSwitch: Where the managed ENI will reside. Ensure that the selected vSwitch has enough available IPs. The ENI will use one IP address from available IPs as the source IP for all probe packets.

     • Security Group: Select a security group for the managed ENI. Ensure that the outbound rules allow ICMP traffic to the destination IP. If the dropdown list is empty, create a security group first.

     • Probe Frequency: The interval at which ping packets are sent. A shorter interval provides more timely connectivity results. Available intervals are 15 Seconds, 1 Minute, 5 Minutes, 15 Minutes, 30 Minutes, and 60 Minutes.

3. Click Create. The probe task starts automatically.

Compare probe tasks

Compare key metrics from multiple probe tasks to quickly spot differences in availability, response time, and packet loss. Use this to identify network issues or compare performance across regions and ISPs.

What to watch for:

• Low availability: If a probe task shows significantly lower availability or a downward trend, check destination service status and network connections.

• High response time: If the response time spikes or remains high, investigate server load, network bottlenecks, or DNS resolution delays.

• High packet loss: If the packet loss rate spikes, check network connection quality and intermediate nodes.

To compare tasks:

1. Click the Task Name/ID of a probe task to go to its details page, and click Compare Tasks in the upper-right corner.

2. Select the Task to compare and choose a time range. Supported comparison metrics include Availability, Response Time, and Packet Loss Rate. You can also select a different Aggregation Cycle.

   The aggregation cycle is the interval for collecting and summarizing data. A shorter period provides finer-grained data, while a longer period is better for viewing overall trends.

Configuring alerts

Add alert rules for any probe task you have created. If a monitored metric reaches a threshold, the system automatically sends a notification by SMS or email.

1. Go to the Site Monitoring page in the CloudMonitor console. In the Actions column for the target probe task, click Edit.

2. In the left-side navigation pane, click Set Alarm. Then, Configure Alerm Rules and Configure Alert Contact Group. For details, see the CloudMonitor documentation:

   • Create an alert rule

   • Create an alert contact or alert contact group

Billing

Traffic Probe is a feature of CloudMonitor and is billed accordingly. For more information, see Site Monitoring Billing.