Category: Virtual Environment

Virtual Machines Hypervisors : VMWare ESXi, Proxmox VE

Remove No Subscription message in Proxmox VE 9

Damien 
#!/bin/sh
WEB_JS=/usr/share/javascript/proxmox-widget-toolkit/proxmoxlib.js
if [ -s "$WEB_JS" ] && ! grep -q NoMoreNagging "$WEB_JS"; then
    echo "Patching Proxmox VE Web UI nag..."
    sed -i -e "/data\.status/ s/!//" -e "/data\.status/ s/active/NoMoreNagging/" "$WEB_JS"
fi
echo "Restarting Proxmox VE Web UI..."
systemctl restart pveproxy.service

This has been tested in Proxmox VE 9.1 and must be run after each update of the Proxmox User Interface. (source : https://community-scripts.github.io/ProxmoxVE/)

In the file : 

/usr/share/javascript/proxmox-widget-toolkit/proxmoxlib.js

search for the line : 

.data.status.toLowerCase() !== 'active') {

and replace with :

.data.status.toLowerCase() == 'active') {

Then, restart the Web UI : 

systemctl restart pveproxy.service

Mount VMWare VMFS Volumes in Linux

Damien

Install the vmfs pckages : 

sudo apt install vmfs6-tools

Mount the partition sdX1 using FUSE : 

vmfs6-fuse /dev/sdX1 /mnt/media

About VMWare VMFS’ hidden files

There are hidden files in the root of any VMFS volume which start with a dot and end with the extension .sf, and some free space seems to be missing from a newly created VMFS datastore.

These system files are indexes and descriptors of the files on the VMFS filesystem and cannot be deleted.

  • They may take more space if you have a small block size (more addresses).
  • They may take less space if you have a large block size (less addresses).
  • They are metadata files that are created when the volume is first set up.
  • They cannot be changed and cannot be deleted.
  • They may grow over time but not significantly.

The VMFS-3 file system organizes all space on disk in collections of resources. There are four resource types managed by the VMFS-3 file system: Blocks, sub-blocks, pointer blocks, and file descriptors.

Resources are grouped together into collections called CLUSTERs and clusters can be further grouped together into CLUSTER GROUPS.

Each resource type is managed by one of several system files which are created during the file system bootstrap process:

  • fbb.sf – file block bitmap.system file
  • fdc.sf – file descriptor cluster.system file
  • pbc.sf – pointer block cluster.system file
  • sbc.sf – sub-block cluster.system file
  • vh.sf – vmfs heartbeat.system file
  • sdd.sf – scsi device description.system file

The VMFS-5 file system uses one additional system file:

  • pb2.sf – pointer block 2.system file

Passing a Physical Disk Drive to a Proxmox VM

Damien

In the Proxmox VE host, display the local disks in the Disks section and note the device you’ll be passing to your VM (e.g. /dev/sdX).

Then, run the host console in the web UI or connect via ssh.

Take note of the VM ID that the physical hard disk will be connecting to.

list the local disks available (since we’ll be passing a physical disk, ignore partitions) : 

ls -n /dev/disk/by-id/

sample output : 
lrwxrwxrwx 1 0 0  9 Feb 11 10:05 ata-CT1050MX30SSD1_1723175912B9 -> ../../sdb
…
lrwxrwxrwx 1 0 0  9 Feb 11 10:05 ata-CT525MX300SSD1_171816E67E70 -> ../../sdc
…
lrwxrwxrwx 1 0 0  9 Feb 11 10:05 ata-CT500MX500SSD1_21102D720097 -> ../../sda

Make sure the VM is stopped before changing its configuration.

Using the VM ID noted earlier, pass the disk to your VM either as a virtio or sata (recommended) device : 

qm set VMID -sata1 /dev/disk/by-id/YYY

usage:
qm set %VMID% -%VIRTUALDEVICE% %DISKID

%VIRTUALDEVICE% can ben set as sataN where N is a digit between 0 and 5.

Then, check in the VM’s Hardware tab that the disk is listed.

Verify inside the guest OS that the physical disk is connected, then it can then be used normally as a physical disk (mount partitions, etc…)

Creating a Proxmox EL9 VM

Damien

The CPU type must be ‘host’; otherwise the VM will kernel panic while booting the installer ISO.

That’s because EL9 now requires CPUs to be compatible with the x86-64 v2 instruction set, which the default Qemu x64 exposed vCPU doesn’t support.

Add TLS Certificate to ESXi

Damien

1. Allow SSH access to ESXi/ESX hosts with public/private key authentication

To allow root access on the ESXi host, change PermitRootLogin no to PermitRootLogin yes in the /etc/ssh/sshd_config file. To disable password login, ensure that the ChallengeResponseAuthentication and PasswordAuthentication are set to no.

If changes were made to the SSHd config file, reload the service :

For ESXi :
/etc/init.d/SSH restart

For ESX :
service sshd reload

After activating the SSH service on the ESXi host (in the web UI : Host tab > Actions menu > Services > Enable Secure Shell (SSH)), backup the original certificates : 

cd /etc/vmware/ssl
mv rui.key rui.key.ori
mv rui.crt rui.crt.ori

For ESXi 5.x, 6.0, 6.5 and 6.7, the authorized_keys is located at :

/etc/ssh/keys-<username>/authorized_keys

Copy the generated public/private key and test the connection to the ESXi host.

(source : https://kb.vmware.com/s/article/1002866)

2. Deploy TLS Certificates to ESXi

Create a TLS certificate (e.g. via acme.sh) and copy to the ESXi host the newly generated certificate : 

scp tls_esxihost.cer root@esxihost.lan:/etc/vmware/ssl/rui.crt 
scp tls_esxihost.key root@esxihost.lan:/etc/vmware/ssl/rui.key

Then, reload services to serve the new certificate : 

services.sh restart

(source : https://www.linuxtricks.fr/wiki/wiki.php?title=vmware-esxi-ajouter-son-propre-certificat)