Setup PXE Network Boot Server in CentOS/RHEL

PXE Server – PXE, an abbreviation of the preboot execution environment, allows us to deploy operating systems on multiple systems automatically at a time in the network. Also, the PXE server helps you to install an OS in a remote system that doesn’t have many options for CD/DVD or USB drives. Now, let start how to setup PXE Network Boot server in CentOS.

DNSMASQ provides DNS and DHCP services, Syslinux package which provides bootloaders for network booting, TFTP-Server, which makes bootable images available to be downloaded via the network using Trivial File Transfer Protocol (TFTP) and VSFTPD Server which will host the local mounted mirrored DVD image

Install DnsMasq Server and configure

use the following command to install DNSMASQ daemon.

# yum install dnsmasq

DNSMASQ main default configuration file located in /etc directory is self-explanatory but intends to be quite difficult to edit, do to its highly commented explanations.

First make sure you backup this file in case you need to review it later and, then, create a new blank configuration file using your favorite text editor by issuing the following commands.

# mv /etc/dnsmasq.conf  /etc/

# vim /etc/dnsmasq.conf

Copy and paste the following example configurations on dnsmasq.conf file and make sure your network settings accordingly.

# DHCP range-leases
dhcp-range= ens192,,,,1h
# Gateway
# Broadcast Address
# NTP Server

pxe-prompt=”Press F8 for menu.”, 60
pxe-service=x86PC, “Install CentOS 7 from network server”, pxelinux

The statements that you need to change are followed:

interface – Interfaces that the server should listen and provide services.
bind-interfaces – Uncomment to bind only on this interface.
domain – Replace it with your domain name.
dhcp-range – Replace it with IP range defined by your network mask on this segment.
dhcp-boot – Replace the IP statement with your interface IP Address.
dhcp-option=3, – Replace the IP Address with your network segment Gateway.
dhcp-option=6, – Replace the IP Address with your DNS Server IP – several DNS IPs can be defined.
server= – Put your DNS forwarders IPs Addresses.
dhcp-option=28, – Replace the IP Address with network broadcast address –optionally.
dhcp-option=42, – Put your network time servers – optionally ( Address is for self-reference).
pxe-prompt – Leave it as default – means to hit F8 key for entering menu 60 with seconds wait time..
pxe=service – Use x86PC for 32-bit/64-bit architectures and enter a menu description prompt under string quotes. Other values types can be: PC98, IA64_EFI, Alpha, Arc_x86, Intel_Lean_Client, IA32_EFI, BC_EFI, Xscale_EFI and X86-64_EFI.
enable-tftp – Enables the build-in TFTP server.
tftp-root – Use /var/lib/tftpboot – the location for all netbooting files.

SYSLINUX Bootloaders Installation

Syslinux is a collection of boot loaders capable of booting from drives, CDs, and over the network via PXE.

# yum install syslinux

The PXE bootloaders files reside in /usr/share/syslinux absolute system path, so you can check it by listing this path content.

# ls /usr/share/syslinux

TFTP-Server Installation

Trivial File Transfer Protocol (TFTP) is a simple lockstep File Transfer Protocol that allows a client to get a file from or put a file onto a remote host. One of its primary uses is in the early stages of nodes PXEBOOT from a local area network.

# yum install tftp-server

Now, copy all SYSLINUX bootloders packages from the above listed location to /var/lib/tftpboot

# cp -r /usr/share/syslinux/* /var/lib/tftpboot

PXE Server Configuration Setup

The PXE Server reads its configuration from a group of specific files (GUID files – first, MAC files – next, Default file – last) hosted in a folder called pxelinux.cfg.

Create a directory pxelinux.cfg and update the default file with below example config.

# mkdir /var/lib/tftpboot/pxelinux.cfg

# touch /var/lib/tftpboot/pxelinux.cfg/default

# vim /var/lib/tftpboot/pxelinux.cfg/default

default menu.c32
prompt 0
timeout 300

menu title ########## PXE Boot Menu ##########

label 1
menu label ^1) Install CentOS 7 x64 with Local Repo
kernel centos7/vmlinuz
append initrd=centos7/initrd.img method= devfs=nomount

label 2
menu label ^2) Install CentOS 7 x64 with Repo
kernel centos7/vmlinuz
append initrd=centos7/initrd.img method= devfs=nomount ip=dhcp

label 3
menu label ^4) Install ESXI6 with local repo
kernel esxi650/mboot.c32
append  esxi650/boot.cfg method= devfs=nomount

label 4
menu label ^4) Boot from local drive

Update CentOS Boot Images to PXE Server

It would require to copy CentOS kernel and initrd files from the ISO. You can directly attach ISO CD and copy the files or download the ISO file to your local machine and mount to copy the files.

# mount -o loop /dev/cdrom  /mnt

# ls /mnt

Download ISO to your local and copy the files.

# cd /opt/

# wget

# mount -o loop /opt/CentOS-7-x86_64-DVD-1511.iso  /mnt

Now, copy CentOS bootable kernel and initrd images to the tftpboot directory

# mkdir /var/lib/tftpboot/centos

# cp /mnt/images/pxeboot/vmlinuz  /var/lib/tftpboot/centos

# cp /mnt/images/pxeboot/initrd.img  /var/lib/tftpboot/centos

Setup FTP Local Mirror to keep the image sources

This is optional setup to boot images from your local FTP mirrors. It is also possible to configure VNC server on CentOS and configure the PXE boot.

# yum install vsftpd

# cp -r /mnt/*  /var/ftp/pub/ 

# chmod -R 755 /var/ftp/pub

Unmount once you’re copied.

# umount /mnt

Start services

# systemctl start dnsmasq
# systemctl status dnsmasq
# systemctl start vsftpd
# systemctl status vsftpd
# systemctl enable dnsmasq
# systemctl enable vsftpd

Verify FTP local mirror network path in your browser. If it is not accessible, make sure port has enabled in your firewall.

Enable firewall ports

# firewall-cmd –add-service=ftp –permanent  ## Port 21
# firewall-cmd –add-service=dns –permanent  ## Port 53
# firewall-cmd –add-service=dhcp –permanent  ## Port 67
# firewall-cmd –add-port=69/udp –permanent  ## Port for TFTP
# firewall-cmd –add-port=4011/udp –permanent   ## Port for ProxyDHCP
# firewall-cmd –reload  ## Apply rules

Verify the PXE boot from your client

Make sure your client machine has enabled in BIOS as Network Boot as primary boot device. In order to choose network booting. After first PXE prompt appears, press F8 key to enter presentation and then hit Enter key to proceed forward to PXE menu.

Once you have reached PXE menu,  hit Enter key and continue with the installation

Now, select your variant and hit enter to continue the installation. That’s it.


Create NFS datastore in VMware Web Client

ESXi can access a designated NFS volume, mount the volume, and use it for its storage needs. You can use NFS volumes to store and boot virtual machines in the same way that you use VMFS datastores.

ESXi supports the following shared storage capabilities on NFS volumes:

  • vMotion
  • VMware DRS and HA
  • ISO image CD-ROMs to virtual machines
  • Virtual Machine snapshots

ESXi does not impose any limits on the NFS datastore size.

Create NFS datastore on your VMware vCenter

  • Login your Web Client
  • Click “Hosts and Clusters” under vCenter
  • Click -> your host -> Datastores – > “New Datastores

  • Select NFS on “Type” -> Click Next
  • Select NFS Version (NFS 3) -> Click Next
  • Update your Datastore Name, Folder location and Server details -> Click Next
  • Finish

If you want refresh your datastore and make sure you connected with NFS storage.

Mount NFS Datastore to Multiple Hosts

Starting from vSphere 6.0 VMware introduced really handy feature to mount NFS Datastore to additional hosts directly from Web Client. It will be very useful to use vMotion and other above feature when you add multiple hosts.

Steps to add multiple hosts in existing NFS datastore.

  • Choose datastore which you would like to add to additional Hosts
  • Right Click or Click Actions

  • Click “Mount Datastore to Additional Hosts
  • Select the hosts that you want to mount the datastore on.
  • Click  Ok

It will mount automatically to all the additional ESXi host. NFS 3 and NFS 4.1, ESXi supports the AUTH_SYS security, also NFS 3 with ESXi does not provide multipathing support. When you create an NFS datastore, make sure all the security and Network connections.


Nginx Basic auth for private registry pull and push

When I was looking for a way to tell the Nginx to let all GET requests through the separate user list and all other requests (e.g. POST, PUT, etc) should be authenticated via the different user lists. This Nginx configuration allows restricting access via different methods to separate users.

This is very useful for the private docker registries, where you want every member of your team to be able to fetch Docker images, but only some users example admin, super admin to push new images to the registry.


  • User write team can use GET, POST, PUT, DELETE and everything else.
  • User read team can only use GET and HEAD.
  • Anonymous users are denied access entirely.



We can use limit_except to configure this setup in Nginx.

Limits allowed HTTP methods inside a location.

Syntax: limit_except method ... { ... }

The method parameter can be one of the following: GET, HEAD, POST, PUT, DELETE, MKCOL, COPY, MOVE, OPTIONS, PROPFIND, PROPPATCH, LOCK, UNLOCK, or PATCH. Allowing the GET method makes the HEAD method also allowed.

Configure the Nginx authentication for the docker private registry pull user accounts and push user accounts using limit_except.

create docker pull users auth file.

# htpasswd -c /etc/nginx/.htpasswd_read read
update you password


create docker pull/Push users auth file.

# htpasswd -c /etc/nginx/.htpasswd_write write
update you password


Add these settings on your nginx config v2 location for the docker registry v2 setup. see the example of how to set up a private Docker registry with Nginx.

# vi nginx.conf

location /v2/ {
auth_basic "read";
auth_basic_user_file /etc/nginx/.htpasswd_read;
limit_except GET {
auth_basic "write";
auth_basic_user_file /etc/nginx/.htpasswd_write;
proxy_pass http://docker-registry;


Now you can test docker registry pull and push using your read, write users.