Issue/Introduction

An abnormal VPC network can have multiple symptoms and causes. Use the corresponding resolution to troubleshoot each scenario.

Environment

• Product: ZCF
• Component: Cloud
• Version: 3.x-5.5.30

Symptoms

An abnormal VPC network generally has one of the following three symptoms:

• Symptom 1: After a host bond is changed, the VPC network used by a VM instance cannot communicate normally.
• Symptom 2: VM instances created on the same VPC network cannot communicate with each other.
• Symptom 3: VM instance networks under different VPC vRouters cannot communicate.

Possible causes

Cause 1

Symptom 1 might occur because the service NIC bridge on the host is deleted while the bond is being changed, preventing the VM instance VPC network from communicating normally.

Cause 2

Symptom 2 might occur because of the VM instance security group configuration.

Cause 3

Symptom 3 might occur because the VPC vRouter does not have a route table attached.

Resolution

Resolution for cause 1

• Restart the affected VM instances and VPC vRouter. This operation creates the bridge automatically.
• Live migrate the affected VM instances and VPC vRouter. This operation creates the bridge automatically, and communication recovers after VM instance live migration.

Resolution for cause 2

• Check the security group rules attached to the VM instances. After confirming that the ingress and egress rules are correct, test again.

Resolution for cause 3

• To enable VPC subnets under two different VPC vRouters to communicate, attach route tables to both vRouters and configure route rules that connect the networks.