How to handle the failure of deploying DevOps tools?

Deployment failure in a dual-NIC cluster with error message “check xx alive failed”

In a dual-NIC cluster, if you deploy DevOps tools from the DevOps Toolchain, the deployment may fail with an error message “Internal error occurred, check xx alive failed: context deadline exceeded”.

Root cause

In a dual-NIC cluster, not all NICs can be used for external access. If you set an internal IP address as the default node IP when creating the cluster and later use it as the access URL for DevOps tool, the tool become inaccessible after deployment.

Solution

1. Log in to the node where the access URL belongs via SSH.

2. Query the default route and associated NIC that can be used for external access.

   # ip r
   default via 192.168.10.10 dev eth0 proto static metric 100 #eth0 is the NIC in question in this example

3. Query the IP address of the NIC you just get.

   # ip a s eth0
   eth0:
   ...
   inet 192.168.200.200/22 #IP address for eth0 in this example

4. On the left navigation bar, click DevOps Toolchain > Integrations.

5. Click the tool name。

6. In the tool details, click the instance name.

7. Click Actions > Update。

8. Update the URL with the above IP address.