Help Center/ Elastic Cloud Server/ User Guide (Ankara Region)/ FAQs/ ECS Failure FAQ/ Why Can't My Linux ECS Obtain Metadata?
Updated on 2024-11-27 GMT+08:00

Why Can't My Linux ECS Obtain Metadata?

Symptom

The security group of the Linux ECS has been configured based on the prerequisites in Obtaining Metadata in the outbound direction, but the ECS still cannot obtain the metadata through the route with the destination of 169.254.169.254.

Root Cause

Run the following command on the Linux ECS configured with a static IP address:

# ip route| grep 169.254

The route with the destination of 169.254.169.254 does not exist, but the route with the destination of 169.254.0.0/16 exists.

Figure 1 Route information

After the network is restarted, the original route with the destination of 169.254.169.254 is changed to the route with the destination of 169.254.0.0/16 without a next hop, as shown in Figure 1. As a result, the Linux ECS cannot obtain metadata.

Solution

  1. Add the route with the destination of 169.254.169.254, and specify the next hop (gateway) and the output device (primary NIC of the Linux ECS). The following is an example:

    # ip route add 169.254.169.254 via 192.168.1.1 dev eth0

    192.168.1.1 is the gateway address of the subnet that the primary NIC resides, and eth0 is the primary NIC.

  2. Run the following command to verify that the metadata can be obtained:

    # curl http://169.254.169.254

    Figure 2 Obtaining metadata
  1. Run the following command to create or modify the /etc/sysconfig/network-scripts/route-eth0 file to prevent the static route from being changed after network restart:

    # vi /etc/sysconfig/network-scripts/route-eth0

    Add the following content to the file:

    In this example, the primary NIC is eth0 and gateway address is 192.168.1.1. Replace them based on site requirements.

    # 169.254.169.254 via 192.168.1.1