<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Homelab &amp; Proxmox on Luther Monson</title>
    <link>https://luth.io/categories/homelab/</link>
    <description>Recent content in Homelab &amp; Proxmox on Luther Monson</description>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Sun, 04 Oct 2026 19:40:00 -0700</lastBuildDate>
    <atom:link href="https://luth.io/categories/homelab/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Breaking Changes Without Breaking Trust: Five Years to go-proxmox v0.8</title>
      <link>https://luth.io/blog/2026/10/go-proxmox-five-years-to-v0-8/</link>
      <pubDate>Sun, 04 Oct 2026 19:40:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/10/go-proxmox-five-years-to-v0-8/</guid>
      <description>&lt;p&gt;&lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt; had its first commit in June 2021 and v0.8.2 went out on&#xA;September 30th. In between there were 366 commits and 43 people besides me with a merged PR. More to the point of this&#xA;post, there are downstream projects that I don&amp;rsquo;t control pinning it in their &lt;code&gt;go.mod&lt;/code&gt;. IONOS&amp;rsquo;s&#xA;&lt;a href=&#34;https://github.com/ionos-cloud/cluster-api-provider-proxmox&#34;&gt;Cluster API provider&lt;/a&gt; and Sidero&amp;rsquo;s&#xA;&lt;a href=&#34;https://github.com/siderolabs/omni-infra-provider-proxmox&#34;&gt;Omni Proxmox provider&lt;/a&gt; are both on v0.8.1 today. A library&#xA;with users like that gets judged less on its API. It gets judged more on what happens to their build when they bump the&#xA;version. I didn&amp;rsquo;t always get that right, so here is the history, measured.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Rescuing a Windows Dynamic Disk Over USB with WSL2 and USB/IP</title>
      <link>https://luth.io/blog/2026/08/windows-dynamic-disk-usb-wsl2-usbip/</link>
      <pubDate>Sat, 01 Aug 2026 13:15:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/08/windows-dynamic-disk-usb-wsl2-usbip/</guid>
      <description>&lt;p&gt;I pulled a 1 TB NVMe out of an old Windows box, dropped it in a USB enclosure and plugged it in. Windows showed nothing.&#xA;No drive letter, no &amp;ldquo;initialize this disk&amp;rdquo; prompt, nothing in Explorer. Disk Management listed it as &lt;code&gt;Invalid&lt;/code&gt;. The&#xA;enclosure&amp;rsquo;s booklet said to &amp;ldquo;bring the disk online&amp;rdquo;, which silently did nothing. The data was completely intact. Windows&#xA;was refusing on &lt;strong&gt;policy&lt;/strong&gt; grounds, not because anything was broken. Getting it back meant going around the Windows&#xA;storage stack entirely. I handed the raw USB device to a Linux kernel that doesn&amp;rsquo;t share Windows&amp;rsquo; opinions.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Five Ways to URL-Encode an SSH Key (and Why Proxmox Rejects Them)</title>
      <link>https://luth.io/blog/2026/07/five-ways-to-url-encode-an-ssh-key/</link>
      <pubDate>Tue, 28 Jul 2026 20:12:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/07/five-ways-to-url-encode-an-ssh-key/</guid>
      <description>&lt;p&gt;&lt;a href=&#34;https://github.com/luthermonson/go-proxmox/issues/144&#34;&gt;Issue #144&lt;/a&gt; on go-proxmox was opened in April 2024 with the&#xA;title &amp;ldquo;Weird sshkeys SSH public key validation error&amp;rdquo;. It is one of my less proud threads. It&amp;rsquo;s thirteen comments of&#xA;people, including me, trying to get a public key into a cloud-init VM through the API. The working answer took more&#xA;than a year to show up. The fix that finally landed in May is one line. Here&amp;rsquo;s why it took so long.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Two Small go-proxmox Fixes: A Watch Channel That Never Closed, and WithRetry</title>
      <link>https://luth.io/blog/2026/07/go-proxmox-watch-close-withretry/</link>
      <pubDate>Sat, 11 Jul 2026 17:26:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/07/go-proxmox-watch-close-withretry/</guid>
      <description>&lt;p&gt;Here are two changes from the v0.5 to v0.7 stretch of &lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt;. They&amp;rsquo;re&#xA;too small for a post each. They&amp;rsquo;re worth knowing about if you write long-running tooling against Proxmox. One is a bug&#xA;in code I wrote about in 2022. The other is a feature that replaces a retry loop everybody was writing for themselves.&lt;/p&gt;&#xA;&lt;h3 id=&#34;the-channel-that-only-closed-when-things-went-well&#34;&gt;The channel that only closed when things went well&lt;/h3&gt;&#xA;&lt;p&gt;In 2022 I wrote up &lt;a href=&#34;https://luth.io/blog/2022/03/proxmox-tasks-upid-wait-go/&#34;&gt;how tasks and &lt;code&gt;Watch&lt;/code&gt; work&lt;/a&gt;. &lt;code&gt;Watch&lt;/code&gt; returns a channel of&#xA;log lines. A goroutine tails the task log until the task stops. Then it closes the channel so you can &lt;code&gt;range&lt;/code&gt; over it.&#xA;Here is the tail loop from that post:&lt;/p&gt;</description>
    </item>
    <item>
      <title>Maintaining an Open-Source Library With AI Agents: AGENTS.md and the Rule They Kept Breaking</title>
      <link>https://luth.io/blog/2026/06/go-proxmox-agents-md/</link>
      <pubDate>Sat, 27 Jun 2026 18:47:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/06/go-proxmox-agents-md/</guid>
      <description>&lt;p&gt;On May 9th I merged &lt;a href=&#34;https://github.com/luthermonson/go-proxmox/pull/264&#34;&gt;#264&lt;/a&gt;. It added an &lt;code&gt;AGENTS.md&lt;/code&gt; to&#xA;&lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt; and a stub &lt;code&gt;CLAUDE.md&lt;/code&gt; that just points at it. The next six&#xA;weeks ran through the v0.8.0 migration guide on June 15th. In that time &lt;code&gt;main&lt;/code&gt; took 77 commits and +31,794 / -1,144&#xA;lines. The repo went from 84 to 186 &lt;code&gt;.go&lt;/code&gt; files. Endpoint coverage went from 30% to&#xA;&lt;a href=&#34;https://luth.io/blog/2026/06/go-proxmox-endpoint-coverage/&#34;&gt;100% of the Proxmox API&lt;/a&gt;. Most of that code was typed by coding agents,&#xA;a lot of it by several at once. This is what that looked like from the maintainer&amp;rsquo;s chair. It&amp;rsquo;s mostly the parts that&#xA;went wrong, because that&amp;rsquo;s where &lt;code&gt;AGENTS.md&lt;/code&gt; got its content.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Proxmox JSON, Part 2: null PIDs, 1.12589990684262e&#43;15 and net10</title>
      <link>https://luth.io/blog/2026/06/proxmox-numbers-null-scientific-notation-net10/</link>
      <pubDate>Tue, 23 Jun 2026 20:31:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/06/proxmox-numbers-null-scientific-notation-net10/</guid>
      <description>&lt;p&gt;In 2023 I wrote about &lt;a href=&#34;https://luth.io/blog/2023/11/proxmox-doesnt-know-what-type-a-number-is/&#34;&gt;why go-proxmox has &lt;code&gt;StringOrInt&lt;/code&gt; and friends&lt;/a&gt;.&#xA;Proxmox sends the same number quoted in one place and bare in another, so the client has types that accept both. That&#xA;post ended on a pretty confident note. Since then three more shapes of &amp;ldquo;the JSON is valid and Go still won&amp;rsquo;t take it&amp;rdquo;&#xA;have come in. One of them is a bug I&amp;rsquo;d already fixed once and then reintroduced myself.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Your Zero Value Just Turned Off My TLS Verification: omitempty, Go Defaults and Proxmox</title>
      <link>https://luth.io/blog/2026/06/omitempty-zero-values-proxmox-defaults/</link>
      <pubDate>Sat, 13 Jun 2026 19:52:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/06/omitempty-zero-values-proxmox-defaults/</guid>
      <description>&lt;p&gt;Every Go developer knows &lt;code&gt;omitempty&lt;/code&gt; drops zero values. What takes longer to sink in is that a plain &lt;code&gt;bool&lt;/code&gt; only has&#xA;two states. A caller filling in an options struct means one of three:&lt;/p&gt;&#xA;&lt;ul&gt;&#xA;&lt;li&gt;&amp;ldquo;set it true&amp;rdquo;&lt;/li&gt;&#xA;&lt;li&gt;&amp;ldquo;set it false&amp;rdquo;&lt;/li&gt;&#xA;&lt;li&gt;&amp;ldquo;I didn&amp;rsquo;t touch this, do whatever the server does by default&amp;rdquo;&lt;/li&gt;&#xA;&lt;/ul&gt;&#xA;&lt;p&gt;With &lt;code&gt;omitempty&lt;/code&gt; you lose &amp;ldquo;false&amp;rdquo;. Without it you lose &amp;ldquo;didn&amp;rsquo;t touch it&amp;rdquo;. In&#xA;&lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt; we managed to ship both bugs, in that order. The second one&#xA;quietly turned off TLS certificate verification for anyone downloading ISOs through the API.&lt;/p&gt;</description>
    </item>
    <item>
      <title>From 52% to 100%: Measuring a Go Client Against the Proxmox API Schema</title>
      <link>https://luth.io/blog/2026/06/go-proxmox-endpoint-coverage/</link>
      <pubDate>Sun, 07 Jun 2026 21:36:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/06/go-proxmox-endpoint-coverage/</guid>
      <description>&lt;p&gt;For five years the honest answer to &amp;ldquo;how much of the Proxmox API does go-proxmox cover?&amp;rdquo; was &amp;ldquo;a lot of it? the parts&#xA;people asked for?&amp;rdquo;. Issues would come in for endpoints that didn&amp;rsquo;t exist in the client. I&amp;rsquo;d have no idea how many more&#xA;were missing. At the end of May I finally built the thing that answers the question. Three days after the first run the&#xA;answer was 100%. The more interesting part is how many times the measuring stick was wrong on the way there.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Cloud-Init ISOs in Pure Go with go-proxmox</title>
      <link>https://luth.io/blog/2026/06/cloud-init-isos-pure-go-proxmox/</link>
      <pubDate>Sat, 06 Jun 2026 15:20:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/06/cloud-init-isos-pure-go-proxmox/</guid>
      <description>&lt;p&gt;Proxmox&amp;rsquo;s built-in cloud-init support is fine for the basics: a user, an SSH key, an IP config. It falls apart the second you&#xA;want your own &lt;code&gt;user-data&lt;/code&gt;, &lt;code&gt;vendor-data&lt;/code&gt; or a real &lt;code&gt;network-config&lt;/code&gt;. The GUI and the &lt;code&gt;ci*&lt;/code&gt; options only expose a handful of&#xA;fields. The escape hatch is &lt;code&gt;cicustom&lt;/code&gt; pointing at snippet files, and that means getting files onto the Proxmox host&amp;rsquo;s&#xA;filesystem first. The usual alternative is building a nocloud ISO yourself with &lt;code&gt;genisoimage&lt;/code&gt; or &lt;code&gt;mkisofs&lt;/code&gt;. That means a&#xA;system dependency on whatever machine runs your automation. You get a different package name on every OS and a subprocess to&#xA;babysit.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Staying Logged In to Proxmox: Ticket Renewal, Expiry and Where API Tokens Stop</title>
      <link>https://luth.io/blog/2026/05/staying-logged-in-to-proxmox/</link>
      <pubDate>Sun, 31 May 2026 20:24:00 -0700</pubDate>
      <guid>https://luth.io/blog/2026/05/staying-logged-in-to-proxmox/</guid>
      <description>&lt;p&gt;Back in 2023 I wrote up &lt;a href=&#34;https://luth.io/blog/2023/08/proxmox-api-auth-tickets-tokens-go/&#34;&gt;how go-proxmox logs in&lt;/a&gt; and ended it with a&#xA;shrug. Once a session existed the client never logged in again. A ticket-based client died at the two hour mark. My&#xA;advice was &amp;ldquo;use an API token for anything long running&amp;rdquo;. It took until this month to fix that properly. It took three&#xA;PRs from two of us to get there. Here&amp;rsquo;s the whole arc, plus the one place where an API token isn&amp;rsquo;t an option at all.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Proxmox Doesn&#39;t Know What Type a Number Is</title>
      <link>https://luth.io/blog/2023/11/proxmox-doesnt-know-what-type-a-number-is/</link>
      <pubDate>Sat, 25 Nov 2023 20:05:00 -0700</pubDate>
      <guid>https://luth.io/blog/2023/11/proxmox-doesnt-know-what-type-a-number-is/</guid>
      <description>&lt;p&gt;Every couple of months somebody opens an issue on &lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt; with a&#xA;title that starts with &lt;code&gt;json: cannot unmarshal string into Go struct field&lt;/code&gt;. The most recent one was fixed last week.&#xA;It&amp;rsquo;s the same root cause every time. The Proxmox VE API is backed by Perl, and Perl doesn&amp;rsquo;t really care whether a&#xA;scalar is a number or a string. The JSON encoder on the other end writes whatever the scalar happens to be at the time.&#xA;So a field that is documented as an integer comes back as &lt;code&gt;2048&lt;/code&gt; from one endpoint and &lt;code&gt;&amp;quot;2048&amp;quot;&lt;/code&gt; from another.&#xA;Sometimes it changes on the same endpoint depending on how the value got into the config.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Logging In to Proxmox From Go: Tickets, CSRF Tokens and API Tokens</title>
      <link>https://luth.io/blog/2023/08/proxmox-api-auth-tickets-tokens-go/</link>
      <pubDate>Sun, 27 Aug 2023 19:18:00 -0700</pubDate>
      <guid>https://luth.io/blog/2023/08/proxmox-api-auth-tickets-tokens-go/</guid>
      <description>&lt;p&gt;I just spent a week of evenings on the access side of &lt;a href=&#34;https://github.com/luthermonson/go-proxmox&#34;&gt;go-proxmox&lt;/a&gt;.&#xA;Realms, groups, users and roles all landed. Before any of that, I reworked how a client gets its credentials in the&#xA;first place. Proxmox gives you two different ways to authenticate. They behave differently enough that it&amp;rsquo;s worth&#xA;writing down how each one works on the wire and what the client does with it.&lt;/p&gt;&#xA;&lt;p&gt;&lt;img src=&#34;https://luth.io/img/2023/08/proxmox-api-auth-tickets-tokens-go-flow.svg&#34; alt=&#34;Ticket and API token flows&#34;&gt;&lt;/p&gt;</description>
    </item>
    <item>
      <title>Waiting on Proxmox: UPIDs, Tasks and WaitFor in Go</title>
      <link>https://luth.io/blog/2022/03/proxmox-tasks-upid-wait-go/</link>
      <pubDate>Sun, 20 Mar 2022 16:42:00 -0700</pubDate>
      <guid>https://luth.io/blog/2022/03/proxmox-tasks-upid-wait-go/</guid>
      <description>&lt;p&gt;If you start automating &lt;a href=&#34;https://www.proxmox.com/en/&#34;&gt;Proxmox VE&lt;/a&gt; through the API, one thing trips you up first.&#xA;Almost nothing you ask it to do is finished when the HTTP call returns. Start a VM, clone a template, download an ISO&#xA;to storage, delete a disk&amp;hellip; the response is a single string that looks like a typo. The actual work is happening in a&#xA;worker process on the node. If your next call depends on that work being done, you have to go find out when it&amp;rsquo;s done.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Recreating Proxmox Ve Xtermjs in Go</title>
      <link>https://luth.io/blog/2022/02/recreating-proxmox-ve-xtermjs-in-go/</link>
      <pubDate>Sun, 06 Feb 2022 21:54:53 -0700</pubDate>
      <guid>https://luth.io/blog/2022/02/recreating-proxmox-ve-xtermjs-in-go/</guid>
      <description>&lt;p&gt;Part of a hobby of mine is running a home lab and using on &lt;a href=&#34;https://www.proxmox.com/en/&#34;&gt;Proxmox VE&lt;/a&gt; to do the virtualization.&#xA;Since I work in the space I&amp;rsquo;m slowly trying to build tooling around Proxmox that will let me use the home lab full time for work.&#xA;Usually my work involves building Kubernetes management tooling. You can imagine waiting for public cloud APIs all day&#xA;can take away hours of your day. That&amp;rsquo;s at cost to your employer for both your time and the compute resources.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
