Environment

• Product: ZCF
• Component: ZCE-X
• Affected objects: ZStack 5.4.7, hyperconvergence, four servers, and management network br_bond0
• Version: 6.4.200.1
• Trigger condition: Change the cloud platform management gateway from 10.10.11.1 to 10.10.11.100 without changing the management IP addresses or subnet.

Symptoms

• The management gateway must be changed offline on four servers.
• Only the management gateway changes; the management IP addresses and subnet masks remain unchanged.
• SSH disconnects during the change, so use a directly connected monitor or IPMI.

Diagnosis

1. Confirm that the environment is within scope and record versions, system architecture, operating system, deployment mode, and storage type.
2. Confirm that only the gateway changes, from 10.10.11.1 to 10.10.11.100.
3. Check platform health, prepare a database backup, and confirm that global high availability can be disabled.

Cause

The management gateway on four servers must be changed offline and the platform configuration synchronized to meet the new network plan.

Resolution

1. Prerequisites

• Risk level: Medium
• Perform the change during a maintenance window and disable global high availability in advance.
• Because SSH disconnects, use a directly connected monitor or IPMI.
• Confirm that the new gateway is already configured on the network side.

2. Procedure

1. Check environment health.

• Expected result: The platform has no abnormalities.

2. Disable global cloud platform high availability and back up the database.

• Expected result: High availability is disabled and the backup is complete.

3. On each node, edit the gateway in the physical machine management NIC configuration and reload the network.

   vim /etc/sysconfig/network-scripts/ifcfg-br_bond0
   ifup br_bond0
   route -ne

• Configuration example:

   DEVICE=br_bond0
   NAME=br_bond0
   TYPE=Bridge
   ONBOOT=yes
   BOOTPROTO=static
   IPV6INIT=no
   IPV6_AUTOCONF=no
   DELAY=5
   STP=no
   IPADDR=xx.xx.xx.xx
   NETMASK=255.255.255.0
   GATEWAY=xx.xx.xx.xx

• Expected result: Each node uses the new gateway, as confirmed by route -ne.

4. After changing all nodes, synchronize the management gateway configuration with zsha2.

   zsha2 change-ip -db-root-pw zstack.mysql.password -gateway XXX -myip XXXX -nic br_XXX -slave root:XXXX@XXXX -vip XX
   zsha2 start-node

• Expected result: The platform management gateway configuration is synchronized and node services return to normal.

5. Enable global high availability.

• Expected result: High availability and normal platform operation are restored.

Verification

• Run route -ne and confirm that the management gateway is updated on every node.
• Confirm normal node communication and management network operation.
• Confirm normal platform resources and successful high-availability restoration after zsha2 change-ip.

Risks and rollback

• Risk: An incorrect gateway may make nodes unreachable.
• Risk: SSH disconnects during the change, requiring console or IPMI access.
• Rollback: Restore the original GATEWAY value, reload the network, and synchronize the zsha2 configuration again.