Disk Group fails to import and multiple entries for each disk appear in "vxdisk list" output.

book

Article ID: 100027365

calendar_today

Updated On:

Description

Error Message

No errors are indicated besides the error indicating Disk Group failed to import.

# vxdg import (disk_group_name)

 

VxVM vxdg ERROR V-5-1-10978 Disk group (disk_group_name): import failed:
No valid disk found containing disk group
 

 

Cause

The /etc/vx/disk.info file has invalid / incorrect content for some dmpnodes.

Example entries similar to the following exist in the file:

INVALID c4t50000123456DCDA0d01 0x4e80ec0 0x2 EMC0_01 EMC 602123
INVALID c4t50000123456DCDA0d02 0x4e80ec8 0x2 EMC0_02 EMC 602123
INVALID c4t50000123456DCDA0d03 0x4e80ed0 0x2 EMC0_03 EMC 602123
INVALID c4t50000123456DCDA0d04 0x4e80ed8 0x2 EMC0_04 EMC 602123
INVALID c4t50000123456DCDA0d05 0x4e80ee0 0x2 EMC0_05 EMC 602123
INVALID c4t50000123456DCDA0d06 0x4e80ee8 0x2 EMC0_06 EMC 602123

Note the "INVALID" entry at the beginning of each line.

 

Resolution

Remove the /etc/vx/disk.info and then restart vxconfigd with the following command:

# vxconfigd -k -x syslog

 

To prevent a recurrence of the issue, it is recommended to stop the eventsource daemon while changing / adding / removing LUNs to the systems.

Prior to making SAN changes to LUN's zoned to the system perform the following commands:

1) Stop the event source daemon

# vxddladm stop eventsource

2) Verify the LUNs are seen correctly in the OS.
 

3) Then restart the event source daemon.


# vxddladm start eventsource
 

4)  If the system is clustered, it is recommended to freeze the service groups currently online in the cluster first.

If the system is not part of a cluster, skip this step.

Where clustered, each service group perform the following:

# hagrp -freeze (group_name) -sys (system_name)  

5) Restart the vxconfigd daemon, which will rescan the available devices and build / update the disk.info file.

# vxconfigd -k -x syslog  
 

6) Where clustered, unfreeze the service groups previously frozen.

If the system is not clustered, skip this step.

# hagrp -unfreeze (group_name) -sys (system_name)

 7) Verify the "vxdisk list" output indicates the expected number of disk devices.

 

Applies To

Where DMP nodename Persistence is set to "yes" in the Volume Manager configuration.

Issue/Introduction

A Disk Group fails to import successfully. Multiple device entries in "vxdisk list" exist for each disk.