Tyring to resize a CFS file system but failed with the following errors "VxVM vxresize ERROR V-5-1-7514 Problem running fsadm command for volume data_cfs, in diskgroup ngpsdg"

book

Article ID: 100000928

calendar_today

Updated On:

Description

Error Message

VxVM vxresize ERROR V-5-1-7514 Problem running fsadm command for volume data_cfs, in diskgroup ngpsdg

Resolution

Trying to resize a CFS file system with the following command:
#/usr/lib/vxvm/bin/vxresize -g ngpsdg data_cfs +150g ngpsdg01 but it fails with the following errors:
VxVMvxresize ERROR V-5-1-7514 Problem running fsadm command for volume data_cfs, in diskgroup ngpsdg
 
 
Whenresizing a CFS file system, there are 2 things which need to beconsidered:
1. which node is the master node
2. which node is where the primary file system located
 
Both conditions must exist on the same host, which means the host which acts as a master node must also be the host where the primary file system is located.
 
So, in this case, the master node is on host 01, whereas the primary file system currently on host 02.
So, the very first thing which needs to be done is the primary file system role needs to be switched to host 01, by running the following from the master node (host01):
 
#/opt/VRTS/bin/fsclustadm setprimary /opt/data/ngps/ngps-data
 
but thiscommand fails with the following errors:
 
UX:vxfsfsclustadm: ERROR: V-3-20308: ioctl failed, Input/output error, error5
 
No errors were log pertaining to this issue on the system log. So, why it fails then?
 
However, when check the conditions of the file system (by using the following command),it appears that the file system need to be 'full' fscked:
 
# echo"8192B.p S"| fsdb -F vxfs /dev/vx/rdsk/ngpsdg/data_cfs| grep flags
flags1 mod 0 clean 3cl
 
"Flags 1"indicates a full file system is required.
 
So, the solutions for this issue will be:
1. unmount this CFS file system from all of the nodes (application need to be stopped first, hence downtime is needed)
2. collect'metasave' for this file system, just in case if we need to recover the filesystem
3. run anormal fsck "fsck -t vxfs /dev/vx/rdsk/ngpsdg/data_cfs"
4. if full fsck required, then run "fsck -t vxfs -o full -y/dev/vx/rdsk/ngpsdg/data_cfs"
5. mount the CFS file system again on all of the nodes
6. perform a resize operations again from the host where the master node and the primary filesystem located
 
 
 
 
 
 
 

 

Issue/Introduction

Tyring to resize a CFS file system but failed with the following errors "VxVM vxresize ERROR V-5-1-7514 Problem running fsadm command for volume data_cfs, in diskgroup ngpsdg"
Powered by Wolken Software