Netboot Requirements

The first chapter is dedicated to laying the ground where we will build upon.
Our intent is coming up with a resilient, scalable, highly available & secure infrastructure,
while possibly reducing at minimum later mangement overhead, and automate most trivial tasks.
Advantages of what we will be doing will become clear as we go continuously re-egnineer,
each time with a new discovery, our infrastructure.

Network booting the our Raspberry Pis, will help up us get rid of time spent physically removing,
re-flashing, and inserting back an empty system, due to frequent data corruption issues,
which SD cards are susceptible of, in fact, such storage media is not meant to read/write continously.

For network boot to operate in our environment, two services need to be configured, DHCP and TFTP.
In our case, we have as DCHP server, a pfSense physical firewall.
Follows the procdeure to configurate DHCP to provide our computing nodes with the necessary data,
which will take a step toward our goal. [ the first of many :) ]

We assume you have a pfSense alredy configured, if not, take a look at this link for how to start with it.

1 Go to Services->DHCP Sever->[Interface] Enable DCHP for the interface
Image title

2 Select the device that we want to enable boot for
enable network-boot feature, the tfpt server feature, and insert the IP address of your future tfpt server.
Insert the filename that will be used to start the bootstrap process (bootcode.bin)
Image title
Image title

3 In our sinology NAS, go in Control Panel -> File services Image title

4 Enable Tftp server, inserting the folder we will use to server the firmware to our devices}
Image title

This confuguration, would have the environment ready for network-booting our cluster.

Comments

Subscribe to our Newsletter