Cluster node fails to join after reboot due to port "b" failure.

book

Article ID: 100003101

calendar_today

Updated On:

Description

Error Message

V-16-1-10181:VxFEN driver not configured. Retrying...

VxVM vxdisk ERROR V-5-1-684 IPC failure: Configuration daemon is not accessible

 

Cause

VxFencing fails to start due to the fact that the vxconfigd daemon is not running.

When the vxconfigd daemon is not running, vxfencing cannot properly determine the status of the fencing coordinator disk group, and therefore cannot start the fencing driver.

This prevents the cluster join operation.

 

Resolution

Freeze the cluster nodes using the below command, prior to attempting the following process.

# hasys -freeze

 

Bring the cluster online by executing the following commands:

1) Start vxconfigd manually

# vxiod set 10# vxconfigd -k -m enable

 

2) Start the vxfencing process "port b"

# vxfenconfig -c

 

3) Check the status of the ports.

An expected sample output:

# gabconfig -a

GAB Port Memberships

===============================================================

Port a gen 5fa03b membership 01

Port b gen 5fa03d membership 01

Port h gen 5fa04e membership ;1

Note the membership on port a and b show two nodes, as expected.

 

4) Start the clustering High Availability Daemon (HAD - port h) on this node.

# hastart

 

5) Verify the cluster is running with:

# gabconfig -a

GAB Port Memberships

===============================================================

Port a gen 5fa03b membership 01

Port b gen 5fa03d membership 01

Port h gen 5fa04e membership 01

Ports a, b, and hshould be listed in the output with full membership.

# hastatus -sum

 

Both nodes should be listed as "RUNNING" at the top of the summary output.

If desired, unfreeze the systems in the cluster:

# hasys -unfreeze

 

NOTE: To prevent recurrence of vxconfigd failing to start on boot, verify the following vxconfigd related files:

# ls -l /etc/vx/volboot

This file (volboot) should exist and be exactly 512 bytes in size.

If the file is missing, perform the following commands to recreate it:

# vxconfigd -k -m disable# vxdctl init # vxconfigd -k -m boot# vxconfigd -k -m enable

# ls -l /etc/vx/reconfig.d/state.d/install-db

This file (install-db) should NOT exist. This file prevents VxVM from starting on boot. Remove the file if it exists.

# rm /etc/vx/reconfig.d/state.d/install-db

 

 

 

 

Issue/Introduction

While VxFencing is in use in a cluster, a cluster node fails to join after reboot, reporting an error for Port "b"

Error Message