Goal

EXPERIMENTATION FOR ALP OS

This wrapper is a parallel virsh command to manage a selected group of Virtual Machine. This provides an easy way to execute the same command on a selected group of Virtual Machine. If you want to manage more than 1 VM you need to script you action and this will be done in a sequential way (most of the time).

Features:

• Launch parallel command on mulitiple VM selected by group
• auto-completion of the virsh domain command
• Reports of error/success per VM
• Interactive mode with intelligent prompt (display group, yaml file and connector)
• Interactive terminal or a one shot command
• colored prompt to easily catch up trouble and syntax
• yaml file group: default is groups.yaml, possible to change to any other file

Interactive command:

See the pvirsh.pod file in man/ directory

Demo

Python requires

• python3-libvirt-python
• python3-PyYAML
• python3-curses (to get completion)

Install

• From source code:

git clone https://github.com/aginies/pvirsh
cd pvirsh
./setup.py install

• Package from OBS

Define Virtual Machine groups

It is possible to select VM directly on the connected hypervisir with select_vm, or you can use a group yaml file which contains list of VM per group. By default the script will use groups.yaml. The yaml file looks like:

suse:
  - sle15sp31$
  - sle15sp4

rhel:
  - rhe
  - fedora

windows:
  - win7
  - win10

• sle15sp31$ : will match exactly this machine name
• sle15sp4 : will match all VM, including sle15sp4*

NOTE: select_vm is only available in interactive mode.

TODO

Probably a lot as this is for testing purpose...

• detect VM state before (useful for some command which requires VM running)
• connect to multiple host?
• ....

Usage

usage: Interactive or Non Interactive command tool to manage multiple VM at the same Time

        Non interactive:
        pvirsh -n -f GROUP.yaml --conn CONNECTOR -g VM_GROUP,VM_GROUP2 -c 'CMD CMD_OPTION'

        Example:
        pvirsh -n --conn local -g suse -c 'domstate --reason'
        
       [-h] [-s] [--conn CONN] [-g GROUP] [-f FILE] [-n] [-c CMD]

optional arguments:
  -h, --help            show this help message and exit

help:
  -s, --showgroup       Show group from VM file content

config:
  --conn CONN           Connect to the hypervisor (local | ssh)
  -g GROUP, --group GROUP
                        Group of VM to use (could be a list separated by ,)
  -f FILE, --file FILE  Group file to use as yaml file

exec:
  -n, --noninter        Launch this tool in non interactive mode
  -c CMD, --cmd CMD     Command to execute on a group of VM

Examples

Get the state of all virtual Machine in suse group:

pvirsh.py -n --conn local -c domstate -g suse,rhel


Connected; Version: 6002000
Multiple group selected
Selected group is suse: ['sle15sp3', 'sle15sp41$', 'sle15sp4-2$']
Selected group is rhel: ['rhe', 'fedora', 'plop']
virsh domstate sle15sp31  shut off Done
virsh domstate sle15sp41  shut off Done
virsh domstate sle15sp4-2  shut off Done
virsh domstate sle15sp32  shut off Done
virsh domstate sle15sp33  shut off Done
virsh domstate sle15sp34  shut off Done
virsh domstate rhel8  shut off Done

Setting hard-limit memory to 1.024GB for all VM in suse group:

pvirsh.py -n --conn local -c "memtune --hard-limit 1000000" -g suse


Connected; Version: 6002000
Selected group is suse: ['sle15sp3', 'sle15sp41$', 'sle15sp4-2$']
virsh memtune sle15sp34 --hard-limit 1000000  Done
virsh memtune sle15sp41 --hard-limit 1000000  Done
virsh memtune sle15sp31 --hard-limit 1000000  Done
virsh memtune sle15sp33 --hard-limit 1000000  Done
virsh memtune sle15sp32 --hard-limit 1000000  Done
virsh memtune sle15sp4-2 --hard-limit 1000000  Done

Adding an RNG device to all VM in suse group:

cat rng.xml 
<rng model="virtio">
  <backend model="random">/dev/urandom</backend>
  <address type="pci" domain="0x0000" bus="0x0a" slot="0x00" function="0x0"/>
</rng>

pvirsh.py -n --conn local -g suse,rhel -c "attach-device --current --file rng.xml"

Connected; Version: 6002000
Multiple group selected
Selected group is suse: ['sle15sp3', 'sle15sp41$', 'sle15sp4-2$']
Selected group is rhel: ['rhe', 'fedora', 'plop']
virsh attach-device sle15sp41 --current --file rng.xml Device attached successfully Done
virsh attach-device sle15sp31 --current --file rng.xml Device attached successfully Done
virsh attach-device sle15sp4-2 --current --file rng.xml Device attached successfully Done
virsh attach-device sle15sp33 --current --file rng.xml Device attached successfully Done
virsh attach-device rhel8 --current --file rng.xml Device attached successfully Done
virsh attach-device sle15sp32 --current --file rng.xml Device attached successfully Done
Command was:virsh attach-device sle15sp34 --current --file rng.xml
ERROR: sle15sp34: b'error: Failed to attach device from rng.xml\nerror: unsupported configuration: a device with the same address already exists \n'