Subdisk disk-02 would overlap subdisk disk-01

Problem

vxresize fails with error message "VxVM vxassist ERROR V-5-1-10127 creating subdisk disk-02: Subdisk disk-02 would overlap subdisk disk-01"
Solution
Issue:
Command vxresize fails with an overlap message when there is enough free space in the disk group to grow the volume.


Conditions:
There is enough free space in the disk group to increase the volume size:

# vxassist -g <diskgroup> maxgrow <volume>

Volume <volume> can be extended by <maxgrow_value> to: <maxsize> (<maxsize_MB>Mb)


The vxresize command fails with the following error message:

# vxresize -F vxfs –g <diskgroup> <volume> +1020mb

VxVM vxassist ERROR V-5-1-10127 creating subdisk< diskname>-02:
Subdisk< diskname>-02 would overlap subdisk <diskname>-01

Cause:
The in-memory configuration that tracks free space on disks has become corrupted and is incorrectly reporting a disk as having free space.

Verify this case using the following procedure (replace <diskname> with the text shown in the vxresize error message):

# vxprint -g <diskgroup> -ht | grep <diskname>


# vxprint -ht | grep sdc

dm sdc sdc auto 2048 60126464 –
sd sdc-01 tvol-01 sdc 0 60125184 0 sdc ENA


The line showing 'sdc-01' indicates that the subdisk takes almost the entire available space (60126464 available, 60125184 allocated).

Now examine the vxdg free output -- this output is generated by the in-memory configuration database and not determined by examining the disk:

# vxdg free | grep < diskname>

testdg sdc sdc sdc 0 60125184 –

The vxdg free output indicates that this disk is registered in the in-memory configuration as having a large amount of free space.


Resolution:

This is a known issue and fixed in following patches.

Solaris : 4.1MP2RP3 / 5.0MP1RP4


Workaround:

This issue can be immediately corrected by restarting the vxconfigd daemon. See the following for details on how to perform that operation based on your environment.


For systems running VERITAS Volume Manager in a non-cluster environment
1. Restart vxconfigd:

# /sbin/vxconfigd –k

For systems running VERITAS Volume Manager in a VERITAS Cluster Server environment
1. Freeze all service groups running on the system that have Volume Manager resources:

# hagrp -freeze< service_group_name>

2. Restart vxconfigd:

# /sbin/vxconfigd –k

3. Unfreeze the service groups:

# hagrp -unfreeze <service_group_name>

For systems running Cluster Volume Manager

1. Stop the cluster on the local node:

# hastop –local

2. Restart vxconfigd:

# /sbin/vxconfigd –k

3. Start the cluster on the local node:

# hastart