Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Table of Contents


Warning

It is always advisable to make a backup of the target VM first, ensuring you can recover your original VM should things go wrong !

Introduction

Generally you should only be needing to resize the partiton at the  /data mountpoint on an NMIS VM as this partition contains the following directories:

  • nmis9/
    • database/
    • var/
    • backups
  • omk/
    • var/
  • mongo/
    • ( mongo/ is the directory set as the storage.dbPath set in /etc/mongod.conf )

If it should become necessary to expand the storage space for the partition at the /data mountpoint, the following set of instructions should help you to perform that change with minimal NMIS downtime.

First, determine whether the NMIS VMs' partitions are using Logical Volume Manager (LVM)

First, determine whether the NMIS VM is using LVM:

If the sudo lsblk command produces partition entries in the TYPE column of type lvm, then that partition is using LVM.

Code Block
sudo lsblk

NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sda      8:0    0  120G  0 disk 
├─sda1   8:1    0    1G  0 part /boot
├─sda2   8:2    0   15G  0 part /
├─sda3   8:3    0    1G  0 part [SWAP]
└─sda4   8:4    0  103G  0 part /var
sdb      8:16   0  120G  0 disk 
└─sdb1   8:17   0  120G  0 part /data


In the example command above, using a recent release of the NMIS VM that does not use LVM, we have disks /dev/sda (disk 1) and /dev/sdb (disk 2).

Disk 1 (/dev/sda) has partitions /dev/sda1 to /dev/sda4 and all partitions are TYPE part
Disk 2 (/dev/sdb) has partition /dev/sdb1 of TYPE part

If your NMIS VM is using partition of type lvm for partition at mountpoint /data, then proceed to the paragraph further below NMIS VMs' using Logical Voume Manager (LVM)
Otherwise, continue with the next paragraph NMIS VMs' using Traditional Disk Partitions


NMIS VMs' using Traditional Disk Partitions

First Step, determine the current state

Code Block
df -h

Filesystem      Size  Used Avail Use% Mounted on
...
/dev/sdb1        40G  8.5G   29G  23% /data
...


df -h
command returns this information for /data partition:

Partition:                            /dev/sdb1
Partition Size:                  40G
Partition Used:                   8.5G
Partition Available:           29G
Partition Used %:              23%
Partition Mount Point:  /data

It is recommended that you rerun the df -h command at the end to verify that the resizing has worked.


Second Step,  Resize Storage Hardware

At this point it'll be simplest to open your vmware client, resize the VM's disk 2 to the desired new size and then reboot the VM to make it pick up the size change.

If you would like to avoid any downtime and not reboot, that's also possible if your virtualisation system allows adding or resizing disks "on the go": Virtualbox doesn't allow that, Vmware does. Please note that Vmware doesn't let you resize disks if there are any snapshots present; if that is the case, then these snapshots must be removed before the resize can occur.


Third Step, Informing the OS and resizing the file system

Install growpart:

Code Block
# centos|rhel
sudo yum update
sudo yum install -y cloud-utils-growpart
# debian|ubuntu
sudo apt update
sudo apt install -y cloud-guest-utils


Grow the partition at the /data mountpoint, which we now know from the df -h commands above is /dev/sdb1:

Code Block
# note there is a space between '/dev/sdb' and '1':
sudo growpart /dev/sdb 1

CHANGED: partition=1 start=2048 old: size=83884032 end=83886080 new: size=251656159 end=251658207


Resize the  filesystem, which will take a bit of time and eventuallly tell you that  it has resized the file system for the new extended disk size:

Code Block
# note there is NOT a space between '/dev/sdb' and '1':
sudo resize2fs /dev/sdb1

resize2fs 1.42.9 (28-Dec-2013)
Filesystem at /dev/sdb1 is mounted on /data; on-line resizing required
old_desc_blocks = 5, new_desc_blocks = 15
The filesystem on /dev/sdb1 is now 31457019 blocks long.


Finally check that /dev/sdb1 has been resized as expected:

Code Block
df -h

Filesystem      Size  Used Avail Use% Mounted on
...
/dev/sdb1       118G  8.5G  105G   8% /data
...


Moving and Resizing partitions /dev/sda1 to /dev/sda4 as needed where NMIS VM is using Traditional Partitions

This should not be necessary with regards to the NMIS VM and requires more thought, effort and skill to achieve.

For example, one could consider adding additional disks to the NMIS VM and moving the /home directory and|or any other directory consuming huge disk space to its own mountpoint on the additional disks as an alternative to growing partitions on disk 1 (/dev/sda).

One should also keep in mind that moving or resizing the /boot partition [ fortunately at partition 1 on disk 1 (/dev/sda1) ] can cause the VM not to boot afterwards.

Since disk 1 (/dev/sda) has more than 1 partition, GParted is probably a useful tool for this task:
https://gparted.org/display-doc.php%3Fname%3Dmoving-space-between-partitions


NMIS VMs' using Logical Volume Manager (LVM)

The resizing procedure is quite simple, for size increases at least. The two required steps are:

...

This is possible because the Opmantek Virtual appliance makes use of Linux's excellent LVM support (Logical Volume Management, described in more details here).


First Step, determine the current state

login as as root to the VM and run

...

 /data is the filesystem on top of the logical volume, which has a size of 99GB (plus spare change), of which 39BG are used. It is recommended that you rerun these commands at the end to verify that the resizing has worked.


Second Step,  Resize Storage Hardware

At this point it'll be simplest to open your vmware client, resize the VM's disk 2 to the desired new size and then reboot the VM to make it pick up the size change.

...

In many cases it will be simpler to add another 'disk' to the system rather than resizing either of the existing 'disks'.


Third Step, Informing the OS and resizing the file system

When the VM boots the newly resized disk 2 (aka /dev/sdb) will be detected but volume group and logical volume still need to be told about the change of 'hardware'.

...

Code Block
cat /proc/scsi/scsi
...
Host: scsi2 Channel: 00 Id: 01 Lun:  # the '2' indicates /dev/sdc is the device file
# this marks the whole disk as physical volume
pvcreate /dev/sdc 
# this adddsadds the pv to the volume group
vgextend vg_nmis64_data /dev/sdc

...

should show that /data is now larger than before.