<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: PXE Commercial</title>
    <description>The latest articles on DEV Community by PXE Commercial (@pxecommercialio).</description>
    <link>https://dev.to/pxecommercialio</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4253052%2F8a63b0b6-fb29-44e8-81e3-8cfeb1e99bf5.png</url>
      <title>DEV Community: PXE Commercial</title>
      <link>https://dev.to/pxecommercialio</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/pxecommercialio"/>
    <language>en</language>
    <item>
      <title>PXE provisioning in real datacenters: three failure modes worth designing around</title>
      <dc:creator>PXE Commercial</dc:creator>
      <pubDate>Thu, 19 Nov 2026 05:58:30 +0000</pubDate>
      <link>https://dev.to/pxecommercialio/pxe-provisioning-in-real-datacenters-three-failure-modes-worth-designing-around-4em5</link>
      <guid>https://dev.to/pxecommercialio/pxe-provisioning-in-real-datacenters-three-failure-modes-worth-designing-around-4em5</guid>
      <description>&lt;p&gt;Most PXE tutorials assume a flat lab network: one DHCP server, one TFTP root, MAC addresses you already know, and an image that installs itself. None of those assumptions survive a production rack.&lt;/p&gt;

&lt;p&gt;Here are three failure modes we kept hitting, and the design changes that made them stop hurting.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Blind ARP sniffing to identify the machine that just booted
&lt;/h2&gt;

&lt;p&gt;The classic approach listens for ARP to map IP to MAC, then matches that MAC against the IPMI/BMC entry so the tool can cut power and boot from network.&lt;/p&gt;

&lt;p&gt;It breaks the moment your provisioning network and your IPMI network are physically isolated — which is how most operators run them. And when a heuristic binding guesses wrong, you power-cycle a production host.&lt;/p&gt;

&lt;p&gt;What works better: let iPXE read the SMBIOS fingerprint and drop the host into a &lt;em&gt;pending pool&lt;/em&gt;. Nothing gets power-controlled until a human confirms the physical label, serial or FRU. Read-only until confirmed.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. Windows driver injection, and the password nobody can reset
&lt;/h2&gt;

&lt;p&gt;Linux provisioning is mostly a solved problem: preseed, autoinstall, kickstart. Windows is where unattended builds fall over — a missing storage driver during WinPE setup, a UEFI/GPT partition script that is one character off, or a local admin password nobody wrote down.&lt;/p&gt;

&lt;p&gt;Two changes help:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Keep driver sets attached to the image template, not to the operator's laptop, so a reinstall is reproducible.&lt;/li&gt;
&lt;li&gt;Make offline password recovery a first-class operation. If the only recovery path is "wipe the disk", operators stop using your tooling for anything risky.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  3. Answer files served in cleartext
&lt;/h2&gt;

&lt;p&gt;The answer file usually contains credentials, and it is usually fetched over TFTP on the same LAN everybody can reach. Spoof a MAC and you can pull it.&lt;/p&gt;

&lt;p&gt;Design around it:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;short-lived task tokens, scoped to one machine and one task&lt;/li&gt;
&lt;li&gt;an encrypted template vault rather than plain files on disk&lt;/li&gt;
&lt;li&gt;signing on any command that can cut power or wipe a disk&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  The checklist we ended up with
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Discovery is read-only; power actions need explicit confirmation.&lt;/li&gt;
&lt;li&gt;Verify SMBIOS / serial before any destructive step.&lt;/li&gt;
&lt;li&gt;Cache images locally so a dead uplink does not stop a build.&lt;/li&gt;
&lt;li&gt;Keep task state at the edge — a broken link should not lose a half-built machine.&lt;/li&gt;
&lt;li&gt;Prefer one binary over a five-container stack you have to babysit at 3am.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;None of this is exotic. It is just the difference between a demo and something an operator will trust with a rack.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Disclosure: I work on &lt;a href="https://tntcloud.io" rel="noopener noreferrer"&gt;PXE Commercial&lt;/a&gt;, a self-hosted bare-metal lifecycle automation engine built around these constraints — iPXE discovery, unattended Windows/Linux provisioning, offline recovery, multi-site orchestration, single Go binary. The lessons above apply whichever tool you use.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>devops</category>
    </item>
  </channel>
</rss>
