G020 - K3s cluster setup 03 ~ Debian VM creation.md

G020 - K3s cluster setup 03 ~ Debian VM creation

Your Proxmox VE system is now configured well enough for you to start creating in it the virtual machines you require. In this guide, I'll show you how to create a rather generic VM with Debian. This Debian VM will be the base over which you'll build, in the following guides, a more specialized VM template for your K3s cluster's nodes.

Preparing the Debian ISO image

Obtaining the latest Debian ISO image

First, you'll need to download the Debian ISO image. At the time of writing this, the latest version is 11.1.0 Bullseye.

You'll find the net install version of the ISO for the latest Debian version right at the Debian project's webpage.

After clicking on that big Download button your browser should start downloading the debian-11.1.0-amd64-netinst.iso file automatically.

Storing the ISO image in the Proxmox VE platform

Proxmox VE needs that ISO image saved in the proper storage space it has to be able to use it. This means that you need to upload the debian-11.1.0-amd64-netinst.iso file to the storage you configured to hold such files, the one called hdd_templates.

1. Open the web console and unfold the tree of available storages under your pve node. Click on hdd_templates and you'll reach its Summary page.

   

2. Click on the ISO Images tab, and you'll see the page of available ISO images.

   

   You'll find it empty at this point.

3. Click on the Upload button to raise the dialog below.

   

4. Click on Select File, find and select your debian-11.1.0-amd64-netinst.iso file in your computer, then click on Upload. The same dialog will show the upload progress.

   

5. After the upload is completed, the upload dialog will close on its own and you'll be back to the ISO Images page. Refresh the page, then you'll see the list now updated with your newly uploaded ISO image.

   

   Notice how the ISO is only identified by its file name. Sometimes ISOs don't have detailed names like the one for the Debian distribution, so be sure of giving the ISOs you upload meaningful and unique names so you can tell them apart.

6. Finally, you can check out in the hdd_templates storage's Summary how much space you have left (Usage field).

   

Now you only have one ISO image but, over time, you may accumulate a number of them. So be always mindful of the free space you have left, and remove old images or container templates you're not using anymore.

Building a Debian virtual machine

In this section you'll see how to create a basic and lean Debian VM, then how to turn it into a VM template.

Setting up a new virtual machine

First, you need to configure the VM itself.

1. Click on the Create VM button found at the web console's top right.

   

2. You'll see the Create: Virtual Machine window, opened at the General tab.

   

   From it, only worry about the following parameters:

   • Node: in a cluster with several nodes you would need to choose where to put this VM. Since you only have one standalone node, just leave the default value.

   • VM ID: a numerical identifier for the VM.

     BEWARE!
     Proxmox VE doesn't allow IDs lower than 100.

   • Name: this must be a valid DNS name, like debiantpl or something longer such as debiantpl.your.pve.domain.

     BEWARE!
     The official Proxmox VE says that this name is a free form text string you can use to describe the VM, which contradicts what the web console actually validates as correct.

   • Resource Pool: here you can indicate to which pool of resources (you have none defined at this point) you want to make this VM a member of.

   The only value you need to set here is the name, which in this case could be debiantpl (for Debian Template).

   

3. Click on the Next button and you'll reach the OS tab.

   

   In this form, you only have to choose the Debian ISO image you uploaded before. The Guest OS options are already properly set up for the kind of OS (a Linux distribution) you're going to install in this VM.

   So, be sure of having the Use CD/DVD disc image file option enabled, so you can select the proper Storage (you should only see here the hdd_templates one) and ISO image (you just have one ISO right now).

   

4. The next tab you should go to is System.

   

   Here, only tick the Qemu Agent box and leave the rest with their default values.

   

   The Qemu Agent option "lets Proxmox VE know that it can use its features to show some more information, and complete some actions (for example, shutdown or snapshots) more intelligently".

5. Hit on Next to reach the Disk tab.

   

   In this step you have a form in which you can add several storage drives to your VM, but there are certain parameters that you need to see to create a virtual SSD drive. So, enable the Advanced checkbox at the bottom of this window and you'll see some extra parameters which I've hightlighted in the next snapshot.

   BEWARE!
   The Advanced checkbox affects all the steps of this wizard, but not all of those steps have advanced parameters to show.

   

   From the many parameters showing now in this form, just pay attention to the following ones:

   • Storage: here you must choose on which storage you will place the disk image of this VM. At this point, in the list you'll only see the thinpools you created in the G019 guide.

   • Disk size (GiB): how big you want the main disk for this VM, in gigabytes.

   • Discard: since this drive will be put on a thin provisioned storage, you can enable this option so this drive's image get's shrunk when space is marked as freed after removing files within the VM.

   • SSD emulation: if your storage is on an SSD drive, (like the ssd_disks thinpool), you can turn this on.

   • IO thread: with this option enabled, IO activity will have its own CPU thread in the VM, which could improve the VM's performance and also will change the SCSI Controller of the drive to VirtIO SCSI single.

   BEWARE!
   The Bandwidth tab allows you to adjust the read/write capabilities of this drive, so only adjust them if you're really sure about how to set them up.

   Knowing all that, choose the ssd_disks thinpool as Storage, put a small number as Disk Size (such as 10 GiB), and enable the Discard, SSD emulation and IO thread options.

   

   This way, you've configured the scsi0 drive you see listed in the column at the window's left. If you wante to add more drives, click on the Add button and a new drive will be added with default values.

6. The next tab to fill is CPU. Since you have enabled the Advanced checkbox, you'll see the advanced parameters of this and following steps right away.

   

   The parameters on this tab are very dependant on the real capabilities of your host's CPU. The main parameters you have to care about at this point are:

   • Sockets: a way of saying how many CPUs you want to assign to this VM. Just leave it with the default value (1).

   • Cores: how many cores you want to give to this VM. When unsure on how many to assign, just put 2 here.

     BEWARE!
     Never put here a number greater than the real cores count in your CPU, or Proxmox VE won't start your VM!

   • Type: this indicates the type of CPU you want to emulate, and the closer it is to the real CPU running your system the better. There's a type host which'll make the VM's CPU have exactly the same flags as the real CPU running your Proxmox VE platform, but VMs with such CPU type will only run on CPUs that include the same flags. So, if you migrated such VM to a new Proxmox VE platform that runs on a CPU lacking certain flags expected by the VM, the VM won't run there.

   • Enable NUMA: if your host supports NUMA, enable this option.

   • Extra CPU Flags: flags to enable special CPU options on your VM. Only enable the ones actually available in the CPU Type you chose, otherwise the VM won't run. If you choose the type Host, you can see the flags available in your real CPU in the /proc/cpuinfo file within your Proxmox VE host.

     $ less /proc/cpuinfo

     This file lists the specifications of each core on your CPU. For instance, the first core (called processor in the file) on my Intel Pentium J2900 is detailed as follows.

     processor       : 0
     vendor_id       : GenuineIntel
     cpu family      : 6
     model           : 55
     model name      : Intel(R) Pentium(R) CPU  J2900  @ 2.41GHz
     stepping        : 8
     microcode       : 0x838
     cpu MHz         : 1440.392
     cache size      : 1024 KB
     physical id     : 0
     siblings        : 4
     core id         : 0
     cpu cores       : 4
     apicid          : 0
     initial apicid  : 0
     fpu             : yes
     fpu_exception   : yes
     cpuid level     : 11
     wp              : yes
     flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology tsc_reliable nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 movbe popcnt tsc_deadline_timer rdrand lahf_lm 3dnowprefetch epb pti ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid tsc_adjust smep erms dtherm ida arat md_clear
     vmx flags       : vnmi preemption_timer invvpid ept_x_only flexpriority tsc_offset vtpr mtf vapic ept vpid unrestricted_guest
     bugs            : cpu_meltdown spectre_v1 spectre_v2 mds msbds_only
     bogomips        : 4833.33
     clflush size    : 64
     cache_alignment : 64
     address sizes   : 36 bits physical, 48 bits virtual
     power management:

     The flags available in the CPU are listed on the flags line, and below you can see also the bugs list.

   Being aware of the capabilities of the real CPU used in this guide series, this form could be filled like shown below.

   

7. The next tab is Memory, also with advanced parameters shown.

   

   The parameters to set here are:

   • Memory (MiB): the maximum amount of RAM this VM will be allowed to use.
   • Minimum memory (MiB): the minimum amount of RAM Proxmox VE must guarantee for this VM.

   If the Minimum memory is a different (thus lower) value than the Memory one, Proxmox VE will use Automatic Memory Allocation to dynamically balance the use of the host RAM among the VMs you may have in your system, which you can also configure with the Shares attribute in this form. Better check the documentation to understand how Proxmox VE handles the VMs' memory.

   

   With the arrangement above, the VM will start with 1 GiB and should only be able to take as much as 1.5 GiB from the host's available RAM.

8. The next tab to go to is Network, which shows too its advanced parameters.

   

   Notice how the Bridge parameter is set by default with the vmbr0 Linux bridge. We'll come back to this parameter later but, for now, you don't need to configure anything in this step, the default values are correct for this VM.

9. The last tab you'll reach is Confirm.

   

   Here you'll be able to give a final look to the configuration you've assigned to your new VM before you create it. If you want to readjust something, just click on the proper tab or press Back to reach the step you want to change.

   Also notice the Start after created check. If enabled, it'll make Proxmox VE boot up your new VM right after its creation.

10. Click on Finish and the creation should proceed. Check the Tasks log at the bottom of your web console to see its progress.

    

    Notice how the new VM appear in the Server View tree of your PVE node. Click on it to see its Summary view, as you can also see in the capture above.

The configuration file for the VM is stored at /etc/pve/nodes/pve/qemu-server (notice that it's relative to the pve node) as [VM ID].cfg. So, for this new VM that has the VM ID 100, the file is /etc/pve/nodes/pve/qemu-server/100.cfg and looks like below.

agent: 1
balloon: 1024
boot: order=scsi0;ide2;net0
cores: 2
cpu: host,flags=+md-clear;+spec-ctrl
ide2: hdd_templates:iso/debian-11.1.0-amd64-netinst.iso,media=cdrom
memory: 1512
meta: creation-qemu=6.1.0,ctime=1637241576
name: debiantpl
net0: virtio=CE:AE:AB:C6:8B:7F,bridge=vmbr0,firewall=1
numa: 0
ostype: l26
scsi0: ssd_disks:vm-100-disk-0,discard=on,iothread=1,size=10G,ssd=1
scsihw: virtio-scsi-single
smbios1: uuid=88136df8-90b0-4dc5-9ae7-2b302dd559c9
sockets: 1
vmgenid: 4e4b10c7-f323-48d8-87c7-6a2c4fdc787c

Adding an extra network device to the new VM

Proxmox VE doesn't allow you to assign, in the VM creation wizard, more than one device of any kind. So, if you want to have an extra network device in your VM, you have to do it after you've created the VM in Proxmox VE. And why the extra network device? To allow your future K3s cluster's nodes to communicate directly with each other through the other Linux bridge you already created in the G017 guide.

So, let's add that extra network device to your new VM.

1. Go to the Hardware tab of your new VM.

   

2. Click on Add to see the list of devices you can aggregate to the VM, then choose Network Device.

   

3. You'll see that the raised window is exactly the same as the Network tab you saw while creating the VM.

   

   You have to adjust two fields:

   • Bridge parameter: here you must set the vmbr1 bridge.

   • Firewall checkbox: since this network device is strictly meant for internal networking, you won't need the firewall active here. Disable this checkbox.

   

4. Click on Add and you'll see that the new network device appears immediately as net1 at the end of your VM's hardware list.

   

Installing Debian on the new VM

At this point, your new VM has the minimal virtual hardware setup you need for installing Debian in it.

1. Go back to the Summary tab of your new VM and press the Start button to start the VM up.

   

   Right after it, click on the >_ Console button to raise a noVNC shell window.

2. In the noVCN shell window you should see the following screen.

   

3. In this menu, you must choose the Install option to run the installation of Debian in text mode.

   

4. The next screen will ask you about what language you want to use in the installation process and apply to the installed system.

   

   Just choose whatever suits you and press Enter.

5. The following step is about your geographical location. This will determine your VM's timezone and locale.

   

   Again, highlight whatever suits you and press Enter.

6. If you choose an unexpected combination of language and location, the installer will ask you about what locale to apply on your system.

   

   Oddly enough, it only shows the options you see in the screenshot above so, in case of doubt, just stick with the default United States - en_US.UTF-8 option.

7. Next, you'll have to choose the keyboard configuration that suits you better.

   

8. At this point it will show some progress bars while the installer retrieves components and scans the VM hardware.

   

9. After some seconds you'll end seeing the following screen about the network configuration.

   

   Since this VM has two virtual ethernet network cards, the installer must know which one to use as the primary network device. Leave the default option (the card with the lowest ens## number) and press Enter.

10. Next, you'll see how the installer tries to setup the network in a few progress bars.

    

11. If the previous network autosetup process is successful, you'll end up in the following screen.

    

    In the text box, type in the hostname for this system, bearing in mind that this VM will become just a template to build others. Preferably, input the same name you used previously in the creation of the VM, which in this guide was debiantpl.

12. In the next step you'll have to specify a domain name.

    

    Here use the same one you set for your Proxmox VE system.

13. The following screen is about setting up the password for the root user.

    

    Since this VM is going to be just a template, there's no need for you to type a difficult or long password.

14. The installer will ask you to confirm the root password.

    

15. The next step is about creating a new user that you should use instead of the root one.

    

    In this screen you are expected to type the new user's full name, but since this is going to be your administrative one, input something more generic like System Manager User for instance.

16. The following screen is about typing a username for the new user.

    

    By default, the installer will take the first word you set in the full name and use it (in lowercase) as username. In this guide, this user will be called mgrsys, following the same criteria used for creating the alternative manager user for the Proxmox VE host.

17. On this step you input the password for this new user. Again, since this VM is going to be a template, don't put a complex password here.

    

18. You'll have to confirm the new user's password.

    

19. This step might not show up in your case, but it's just about choosing the right timezone within the country you chose before.

    

20. After seeing a few more loading bars, you'll reach the step regarding the disk partitioning.

    

    Be sure of choosing the SECOND guided option to give the VM's disk a more flexible partition structure with the LVM system, like the one you have in your Proxmox VE host.

    

21. The following screen is about choosing the storage drive to partition.

    

    There's only one disk attached to this VM, so there's no other option but the one shown.

22. Next, the installer will ask you about how to set up the home and other directories.

    

    This VM is going to be a template for servers, so you shouldn't ever need to mount a separate partition for the home directory. Something else could be said about the var or the tmp directories, whose contents can grow notably. But, since you can increase the size of the VM's storage easily from Proxmox VE, just leave the default highlighted option and press Enter.

23. The next screen is the confirmation of the disk partitioning you've setup in the previous steps.

    

    If you are sure that the partition setup is what you wanted, highlight Yes and press Enter.

24. The following step asks you about the size assigned to the LVM group in which the system partitions are going to be created.

    

    Unless you know better, just stick with the default value and move on.

25. After a few loading screens, you'll reach the disk partitioning final confirmation screen.

    

    Choose Yes to allow the installer to finally write the changes to disk.

26. After the partitioning is finished, you'll reach the progress bar of the base system installation.

    

    The installer will take a bit in finishing this task.

27. When the base system is installed, you'll see the following screen.

    

    You don't have any additional media to use, so just answer No to this screen.

28. The following step is about setting the right mirror servers location for your apt configuration.

    

    By default, the option highlighted will be the same country you chose at the beginning of this installation. Chose the country that suits you best.

29. The next screen is about choosing a concrete mirror server for your apt settings.

    

    Stick with the default option, or change it if you identify a better alternative for you in the list.

30. A window will arise asking you to input your HTTP proxy information, if you happen to be connecting through one.

    

    In this guide it's assumed that you're not using a proxy, so that field should be left blank.

31. The installer will take a bit while configuring the VM's apt system.

    

32. When the apt configuration has finished, you'll see the following question about allowing some script to get some usage statistics of packages on your system.

    

    Choose whatever you like here, although bear in mind that the security restrictions that you'll have to apply later to this Debian VM system may end blocking that script's functionality.

33. After another loading bar, you'll reach the Software selection screen.

    

    You'll see that, by default, the installer is setting the VM to be a graphical environment. You only need the two last options enabled (SSH server and standard system utilities), so change them so they look like below and then press Continue.

    

34. Another progress windows will show up and the installer will proceed with the remainder of the installation process.

    

35. In the installation process, the installer will ask you if you want to install the GRUB boot loader in the VM's primary storage drive.

    

    The default Yes option is the correct one, so just press Enter on this screen.

36. Next, the installer will ask you on which drive you want to install GRUB.

    

    Highlight the /dev/sda option and press Enter to continue the process.

    

37. After installing the GRUB boot loader, you'll see for a few seconds how the installer finishes the process.

    

38. After the installation has finished, the installer will warn you about removing the media you used to launch the whole procedure.

    

    BEWARE!
    Keep calm, DON'T press Enter yet and read the following steps.

    If you Continue, the VM will reboot and, if the installer media is still in place, the installer will boot up again.

39. Go back to your Proxmox VE web console, and open the Hardware tab of your VM. Then, choose the CD/DVD Drive item and press the Edit button.

    

40. You'll see the Edit window for the CD/DVD Drive you chose.

    

    Here, choose the Do not use any media option and click on OK.

    

41. Back in the Hardware screen, you'll see how the CD/DVD Drive is now set to none.

    

    BEWARE!
    This doesn't mean that the change has been applied to the still running VM. Usually, changes like these will require a reboot of the VM.

42. Now that the VM's CD/DVD Drive is configured to be empty, you can go back to the noVNC shell and press on Enter to finish the Debian installation. If everything goes as it should, the VM will reboot into the GRUB screen of your newly installed Debian system.

    

43. Press Enter with the default highlighted option or allow the timer to reach 0. Then, after the usual system booting shell output, you should reach the login.

    

44. As a final verification, login either with root or with mgrsys to check that they work as expected.

Note about the VM's Boot Order option

Go to the Options tab of your new Debian VM.

In the capture above, you can see highlighted the Boot Order list currently enabled in this VM. If you press on the Edit button, you'll be able to edit this Boot Order list.

Notice how PVE has already enabled the bootable hardware devices (hard disk, CD/DVD drive and network device) that were configured in the VM creation process. Also see how the network device added later, net1, is NOT enabled by default.

BEWARE!
When you modify the bootable hardware devices of a VM, Proxmox VE WON'T enable automatically any new bootable device in the Boot Order list. You must revise or modify it whenever you make changes to the hardware devices available in a VM.

Relevant system paths

Directories

• /etc/pve
• /etc/pve/nodes/pve/qemu-server

Files

• /etc/pve/user.cfg
• /etc/pve/nodes/pve/qemu-server/100.cfg

References

Debian

• Debian
• How to Install a Debian 10 (Buster) Minimal Server

Virtual Machines on Proxmox VE

• PVE admin guide. Qemu/KVM Virtual Machines
• PVE admin guide. User management. Pools
• PVE admin guide. Qemu/KVM Virtual Machines ~ CPU options
• PVE admin guide. Qemu/KVM Virtual Machines ~ Memory options
• PVE admin guide. Qemu/KVM Virtual Machines ~ Network Device options
• Best practices on setting up Proxmox on a new server

Navigation

<< Previous (G019. K3s cluster setup 02) | +Table Of Contents+ | Next (G021. K3s cluster setup 04) >>