Chameth.com - posts like building-a-new-computer, debugging-beyond-the-debugger, docker-automatic-nginx-proxy, finding-an-awkward-bug-with-claude-code but not g15-ram-upgrade, migrating-from-github-to-forgejo, why-you-should-be-using-https

The Case of the Unchanging Config

2026-04-09T00:00:00Z

Last week I was attempting to make it so I could share pictures on IRC directly from my client. This sounds simple, but it involves a bouncer¹ that proxies the request to a standalone image hosting service that I had to modify to be compatible. At one point my testing loop was:

Make a change to the hosting service
Commit it
Tag a new release
Wait for it to build
Update the version I’m running on my server
Reconfigure the bouncer to pass the new parameter or change the URL or whatever
Try uploading a photo from my phone
Realise I’ve overlooked something and go back to step one

I could have set up a local copy of everything and tested it sensibly. I should have set up a local copy of everything and tested it sensibly. But it seemed like such a trivial change, and setting up the whole environment seemed like such a pain. After the third or so iteration of failure I was pretty annoyed with myself, computers, and basically everything.

My biggest annoyance was that my bouncer would not pick up the new URL from the config when I changed it. It’s meant to reload the config when it receives a SIGHUP, and it claimed to in the logs, but I could clearly see it was still hitting the old URL. Restarting the bouncer to update the config is a pain, as it disconnects me from all the IRC networks, and has to reconnect to them all, reauthenticate, etc. It also mildly spams everyone who shares a channel with me².

When I finally got everything working I had a look at the bouncer source, and thought I’d spotted the issue. I raised a bug report, ending in this remark:

It looks like the config is reloaded properly, but the handler for uploads is created once at startup and has its own copy of the uploader, so effectively snapshots the config to whatever it is at startup:

fileUploadHandler := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
    cfg := srv.Config()
    h := fileupload.Handler{
        Uploader:    cfg.FileUploader,
        DB:          db,
        Auth:        cfg.Auth,
        HTTPOrigins: cfg.HTTPOrigins,
    }
    h.ServeHTTP(w, r)
})

I’m not quite sure what I thought I saw there, but that’s not right. I want to call it a hallucination, but that feels like a very overloaded word these days. I guess I was just seeing what I wanted to see, and that was a good excuse to stop investigating. Still, even if I misidentified the cause, the bug was still present, right? … Right?

The Science of Deduction

When the developer got back and said he couldn’t reproduce my issue, I went back to my install and immediately reproduced it. With a calmer head, I figured it was probably something with my particular set up. First thing to check: am I actually running the version I think I am?

Unfortunately the bouncer doesn’t actually seem to expose the version anywhere that I can see. It’s not in the logs, it doesn’t have a -version flag, and none of the IRC-based status commands seem to include it. But I know it’s a Go app, and I know Go embeds the version information. go version -m <binary> will dump it all out, but the binary is inside a Docker image, and the Docker image is one of my nice, minimal, artisanal ones so doesn’t ship a go binary. No problem, docker compose cp bouncer:/bnc ./bnc yoinks the binary out of the container, and then dumping the version shows that, yes, I am running the version I thought I was. Hmm.

The next tool I reached for in my toolbox was strace. Maybe it’s not actually reading the file for some reason? I immediately executed strace -p <pid> -e trace=openat,open,read,pread64 -f without having to look any part of that up. Yep. Definitely. Then I edited the config, HUP’d the bouncer, and saw that it was… reading the config file. As it’s meant to. By default strace truncates strings to 32 bytes, so I couldn’t actually see the line I’d changed. Some more definitely-not-RTFMing later, and rerunning it with an extra -s 65536 let me see the full config. Surprise! The config hadn’t changed!

To confirm my findings, I used docker compose cp again, this time yoinking the config file from inside the container. The inside config file was definitely different to the outside config file. What? My hypothesis at this point was “something something Docker nonsense”. I mount the config as read-only, and was wondering if that meant that Docker was doing something weird instead of just bind mounting it. A quick trip to /proc/<pid>/mounts showed that it was, in fact, not doing anything weird, and was just bind mounting it.

If the file is bind mounted, then surely it’s the same file? I ran stat on the file on the host, noted the inode number, then pondered how to actually run stat on the file inside the container, given the aforementioned awkwardly minimal image. The solution was easy: access it via /proc/<pid>/root/. I could’ve saved myself a bunch of docker compose cp if I’d thought about that earlier. Oh well. The inode of that file was different. What?

The answer was DNS. It’s always DNS. Oh, sorry, force of habit. I meant the answer was vim. I was editing the config in vim, and when it saves files, by default, it writes the new content to a temporary file and does an atomic rename. That’s normally a good thing: it prevents corruption if the write fails midway through³. In this case, though, that means the new file has a new inode. Bind mounting a file binds to the inode, so the container just keeps clutching onto the original config from when it was started, blissfully unaware that the party has relocated down the street.

This problem is likely to happen whenever you bind mount a file into a container. When I mentioned this to a friend, he immediately responded “oh yeah, never do that”, and went on to describe the horrible hacks he’s had to add to Ansible to sidestep the issue. The nicer solution is to just bind mount an entire directory if you can, as then it doesn’t matter what happens to the files within it. I really like having the config files sat alongside the Docker compose files, though; having to create a directory just to work around some bind mount weirdness upsets me.

Now I knew what the problem was, I found there was an issue raised against Docker twelve years ago⁴. The first response was “that’s expected”, and I fully agree with the author: “respectfully, that might be expected by you, but it was not expected by me”!⁵ As a result of the issue they added a nice note to the docs:

Note: Many tools used to edit files including vi and sed --in-place may result in an inode change. Since Docker v1.1.0, this will produce an error such as “sed: cannot rename ./sedKdJ9Dy: Device or resource busy”. In the case where you want to edit the mounted file, it is often easiest to instead mount the parent directory.

But in the intervening twelve years, both the note and the functionality described have gone missing. Ho-hum.

I found it interesting how I’ve only just hit this problem, given how long I’ve used Docker. But I realised that almost everything I run I’m happy to just restart. Cattle, not pets, and so on. My IRC bouncer is one of the few exceptions to that. The only other thing I regularly hot reloaded was Centauri, my reverse proxy, but that had a whole config directory mounted because it was shared between containers, so nicely sidestepped the foot-gun.

So lesson learnt: check for weird bind mount issues before raising issues about config hot reloading. It’ll join the esteemed company of “maybe the drive is full and causing completely unrelated problems?”, “perhaps everything is dog slow because the kernel ran out of entropy?”, and “did systemd sneakily take over that functionality while you weren’t looking?” in the troubleshooting checklist.

An IRC bouncer is basically an always-on proxy. It connects to the IRC networks for you, then your clients connect to your bouncer. The bouncer can then send incoming messages to all your different clients, cache them when you’re offline and replay them later, and lots of other nice things people take for granted in their chat apps these days. ↩︎
Join/part/quit spam is part of IRC, and clients have ways of handling it, but I still feel bad about doing it excessively. ↩︎
You can disable this in vim by setting backupcopy=yes. Good luck redoing all the debugging if you ever accidentally remove that from your vimrc, though! ↩︎
Which is weird, because Docker can’t possibly be that old. That would make me much older than I’m prepared to accept. ↩︎
If you read the thread they didn’t actually mean to sound so dismissive, but it’s still pretty funny. “Yes, it’s expected that the foot-gun causes your foot to hurt. Duh. What did you expect?” ↩︎

An interesting Tailscale + Docker gotcha

2026-01-29T00:00:00Z

As I’ve written about before, I use Tailscale for a lot of things. I thought I had it set up in a reasonably secure manner, but I recently noticed a problem.

I use Tailscale’s ACLs to limit what each node can access, based on the tags I apply to it. So an app node can’t access anything via Tailscale, while an integration or server node can access things tagged with either app or integration. This is expressed pretty simply in the Tailscale ACL JSON:

{
	"tagOwners": {
		// Servers that can be SSH'd into
		"tag:server": [],
		// Applications that are exposed on tailscale but never connect out
		"tag:app": [],
		// Things which talk to other services over tailscale (connecting to apps etc)
		"tag:integration": [],
	},
	"grants": [
		{
			// Users can access everything
			"src": ["autogroup:member"],
			"dst": ["*"],
			"ip":  ["*"],
		},
		{
			// Servers and integrations can access integrations and apps
			"src": ["tag:server", "tag:integration"],
			"dst": ["tag:integration", "tag:app"],
			"ip":  ["*"],
		},
	],
}

Over the past week I’ve been setting up a private Forgejo instance behind Tailscale, complete with an actions runner that runs things using a docker-in-docker container. I didn’t want the runner knowing anything about Tailscale, so I had it configured to speak to Forgejo direct over HTTP (forgejo:3000) instead of using full Tailscale HTTPS URL that I use when accessing it (http://git.example-net.ts.net/)¹.

Everything was going fine, until I forgot to do that translation… and it worked. My Forgejo action runners could access anything on my tailnet. I run several things on Tailscale that just have authentication turned off, on the basis that only authorised devices can access them. Things like the admin interface for this website. I definitely didn’t intend for any workflow I run on my git server to have access to edit my website!

This wasn’t just limited to Forgejo, either. Any docker container I was running could access the tailnet. In hindsight it’s fairly obvious why: the host is running Tailscale, connected as a node tagged with my server tag. That creates a tailscale0 interface, and automagically sets up iptables rules to route Tailscale traffic over the interface. Docker also automagically sets up iptables rules to bridge traffic, and apparently these two sets of rules interact in such a way that traffic from Docker containers is allowed to route via the tailscale0 interface.

I say it’s fairly obvious in hindsight — there’s no reason why Docker would special case any particular host interface after all — but it still feels pretty surprising. Because both bits of software inject their own iptables rules, I never really had a good mental model for how they interact. The host Tailscale node was a completely separate building block to Docker. It would be a pain to use either of them if they didn’t do these rules, but it’s also one of the reasons I don’t really like “magical” things².

There are lots of ways to fix this, but none of them feel particularly great. You can configure both Tailscale and Docker to not automatically fiddle with iptables and handle the rules yourself, but I really hate dealing with iptables³. As a stopgap I did hold my nose and add some iptables rules to drop traffic to the tailscale0 interface if it originated from the IP ranges that Docker was configured to use:

iptables -I DOCKER-USER -s 192.168.0.0/16 -o tailscale0 -j DROP
iptables -I DOCKER-USER -s 172.17.0.0/12 -o tailscale0 -j DROP

The DOCKER-USER chain is a nice little escape hatch; it comes before the main auto-generated DOCKER chain, and Docker leaves the rules in it alone.

So I added these rules, and felt pretty good about myself, and then… everything started breaking in weird ways. After some debugging I realised the problem was DNS⁴. The server’s DNS resolver is 100.100.100.100, a special Tailscale address. This is so that it can resolve tailnet hostnames via MagicDNS (and so I don’t have to configure my custom DNS servers manually on each device; Tailscale does it for me). My new iptables rules inadvertently dropped all the DNS packets coming from docker containers. D’oh.

Obviously the solution here is to double down and add MORE iptables rules:

iptables -I DOCKER-USER -s 192.168.0.0/16 -d 100.100.100.100 -o tailscale0 -j ACCEPT
iptables -I DOCKER-USER -s 172.17.0.0/12 -d 100.100.100.100 -o tailscale0 -j ACCEPT

This got DNS working again, but it all feels a bit gross. Not least because to actually make these persist you need to use iptables-save, and then you also get all the rules that both Docker and Tailscale have inserted. I came across a script to filter out the Docker ones, but… yuck⁵.

I realised a cleaner idea might just be to lock down what server tagged nodes could access within Tailscale itself. That way I can avoid touching iptables at all. It’s handy to be able to curl services from the host when debugging, but it’s not really necessary. So I removed that access, and… stuff broke again. I’m using my Forgejo instance as a registry for some of the docker images I run, so the docker daemon needs to be able to reach it. I ended up making a new tag for infrastructure, which can be accessed from server devices. This does still allow all the docker containers to reach Forgejo, but I already have it set up with appropriate access controls and public/private repository splits. Forgejo is a service designed to run publicly, so this seems a reasonable trade-off for convenience. I used the tests feature of Tailscale’s ACL config to make sure I’d got the rules right:

{
	"tests": [
		{
			// Servers can only access infrastructure
			"src":   "tag:server",
			"proto": "tcp",
			"allow": [
				"tag:infrastructure:8080",
			],
			"deny": [
				"100.84.16.43:8080",
				"me@example.com:8080",
				"tag:server:8080",
				"tag:app:8080",
				"tag:integration:8080",
			],
		},
	],
}

The whole situation still feels a bit messy. If I ever get around to switching to nftables I might loop back and manually craft some rules for routing traffic, instead of leaving Tailscale and Docker to do their own thing.

I discovered later on that logging into the container registry at forgejo:3000 actually issued a redirect to https://git.example-net.ts.net/ so this was all basically for naught… ↩︎
I didn’t use Tailscale’s MagicDNS for a long time just because the word “magic” put me off. Only when I eventually got around to learning how it worked, and seeing that it wasn’t really that magical under the hood, did I change my mind. ↩︎
nftables seems far better in lots of ways, but I can’t really be bothered migrating. Maybe next time I reimage the server for whatever reason… ↩︎
It’s always DNS… ↩︎
Again, nftables would almost certainly help here. It actually has (gasp) configuration files. But again, I really didn’t want to spend the time migrating. ↩︎

Fixing a loud PSU fan without dying

2025-07-30T00:00:00Z

Three months after I built my new computer, it started annoying me. There would occasionally be a noise that sounded like a fan was catching on a cable, but there weren’t any loose cables to be a problem. Over the course of a few weeks, the sound got progressively worse to the extent that I didn’t want to use the computer without headphones on. I measured the sound at 63 dB, which is about the sound of someone talking. That may not sound terrible, but it’s a constant, nasty noise coming from something that sits about 40cm from my head.

After some investigating¹, I identified the PSU fan as the culprit. I have a Cooler Master V750 SFX, which is not super high-end, but wasn’t cheap, either. It shouldn’t be developing issues after three months. Thankfully, it comes with a ten-year warranty, so it should be easy to get sorted, right?

Warranty woes

I looked at Cooler Master’s warranty, and for issues within the first two years you have to deal with the retailer. That would be Amazon in my case. So I looked at Amazon’s information on warranty issues. Their policy is that if it’s more than 30 days since purchase, you have to send it off to a third-party repair center and wait for them to diagnose and try to repair it. Here’s the kicker:

Usually repairs take up to 20 business days (including delivery time), but could take slightly longer

I use the computer for work, have upcoming LAN parties to go to, and generally can’t do without it for an entire month. That’s assuming they reproduce the issue: the PSU fan only turns on when it reaches a certain temperature, so if you just plug it in briefly it won’t exhibit any symptoms.

If I was going to do a warranty return, I’d need to buy a new PSU first to use while this one was away. If I’m spending £100+ on a PSU, then I’d be getting a better one that doesn’t have known issues with the fans. So when the Cooler Master eventually got returned I’d have no use for it. What a waste.

So the warranty is about as much use as a chocolate teapot. Good to know, I guess. If I wanted decent service, I should have probably used a real retailer, not Amazon. Lesson learnt.

Doing some research

If the warranty is useless to me, maybe I should just void it and try to fix it myself. I did some research online. Most of the results went like this:

if you’re not a trained professional you should not be opening up your PSU at all. There’s a lot of current that goes through it which will seriously hurt/main/kill you

it’s time consuming and potentially dangerous

or some other variety of doom and gloom. There were some actually useful results sprinkled in, though. A post on the smallformfactor.net forums details someone swapping out the fan on the 850W version of the same power supply because it was too noisy. It looks like the fan connector in the PSU is a two-pin connector, and they just used an off-the-shelf adapter to convert the standard four-pin case fan connector. It sounded a bit annoying to make it all fit, though.

A similar post on reddit showed basically the same upgrade, but identified the connector as a 2.54mm JST-XH. They crimped their own connector on, which feels a lot better than using a bulky connector. I do have a crimping set for JST connectors, but I absolutely hate doing them because they’re so fiddly and annoying. I’m not sure if I trust one of my crimps inside a power supply, either.

Looking at the pictures in those posts, I identified the fan as an Ong Hua HA9215VH12FD-F00. I searched to see if I could see any specifications for it. I couldn’t. But what I did get back as my first result was another Reddit post, this one titled “Getting rid of a HA9215SH12FD-F00 fan from your PSU”. It says something about the quality of a component when the first result is a post about getting rid of it because it’s annoying².

All of these mods seemed easy enough, and I’m reasonably confident in my ability to both discharge and avoid touching capacitors, so I probably wouldn’t die. I ordered a Noctua NF-A9x14 fan and some pre-made JST-XH connectors and waited for them to arrive.

Upgrade time

The replacement parts arrived one evening, so when I turned the computer off for the night I unplugged it from the mains, held the power button down for a while, and left it overnight for good measure. The next morning I popped open the case, disconnected and removed the PSU, gleefully tore the “warranty void if removed” sticker off, and opened it up.

Removing the fan was simple: unplug the JST connector and undo the screws holding it to the outer panel. I confirmed that my JST connectors were the right size, then started surgery on the Noctua fan. I carefully removed the heatshrink, cut off the normal fan connector, and separated out the four wires. A quick check of their FAQ confirmed the wiring colours: black was ground, yellow +12V, and blue and green used for speed control (which we’re not using in this instance). I clipped the blue and green wires off near the hub so they wouldn’t be in the way, then cut the yellow and black wires down to length, and attached the JST connector with a solder seal:

Old fan, new fan, burnt mouse mat

In the process I discovered my mouse mat doesn’t like being hit with a heat gun. Whoops. The solder seals are great: the solder in the center melts to form an electrical connection, the parts at the end grip the cable to provide mechanical strength, and the whole thing acts like heatshrink and keeps the connection insulated. A lot less fiddly than crimping JST connectors, and a lot smaller and cheaper than an adapter.

I installed the fan and reassembled the PSU. Then undid it all and tried again, as apparently the face the fan mounts to isn’t quite symmetrical and I’d installed the fan upside down. Connected everything back up, plugged the computer in, booted it up, and… nothing. The PSU doesn’t actually power the fan until it hits a certain temperature. I figured putting it under load would probably heat it up, so ran a graphics benchmark a few times. Eventually the fan spun up. Huzzah.

Soon the entire computer will be beige and brown

Peace at last

I measured the sound of the new fan: 43 dB. That’s a decrease of 20 dB. Due to the weird logarithmic properties of the dB scale that no one understands, that means it sounds about 4 times quieter to human ears. Big improvement!

It’s worth pointing out this isn’t a straight one-to-one upgrade. I don’t know the specs of the old fan in terms of airflow, and I’m just assuming that the Noctua will be good enough by virtue of being a decent fan. They also behave slightly differently as the power ramps up: the stock fan starts spinning at a lower voltage than the Noctua, which needs around 7V to get started properly³. I don’t think this will be an issue, as it’s effectively just putting a small step at the lower end of the fan curve. It might shorten the overall lifespan of the PSU, but from my point of view it failed after four months, so any extra time is a bonus.

I recorded some before/after audio, so you can hear the difference. I think there’s some amount of volume normalisation going on, so it’s not quite as stark a difference as in person, but it’s still noticeable.

Audio sample from the old fan

Audio sample from the new fan

And if you couldn’t hear it, here’s a visual representation:

Side-by-side comparison of the audio clips

The whole upgrade took about an hour (less time than writing this post about it!), and I even managed to avoid killing myself.

Opening the case and sticking my ear near to various components. High-tech stuff. ↩︎
On closer inspection it’s actually a slightly different model — there’s an S not a V in the middle there — but I’m going to assume that doesn’t change much on the quality/noise front. ↩︎
It then keeps spinning down to about 5V or so, it just needs enough power to initially get started. ↩︎

Finding an awkward bug with Claude Code

2025-06-30T00:00:00Z

The many times Claude thought it was right…

I recently encountered a bug in one of my projects that I couldn’t immediately figure out. It was an issue in Centauri, my reverse proxy. After its config was updated, I noticed¹ it stopped serving responses. Looking at the logs, I could see it was obtaining new certificates from Let’s Encrypt for a couple of domains, but I’d designed it so that wouldn’t block requests (or so I thought). After a few minutes of confusion, everything went back to working.

It felt like an issue with locking, but a quick glance didn’t show me anything obvious. I decided to throw the problem at Claude Code and see how it did. With some guidance² it managed to identify the cause, but it was a bit of a journey to get there. I’m going to share the prompts and responses, as it was an interesting process.

The initial prompt

I thought the issue might be quite obvious, and I’d just overlooked it. Centauri isn’t that complex a project, after all. So I gave Claude a fairly minimal prompt to set it off:

> Hello! Please examine this codebase. I have noticed an issue that when a
  certificate is being renewed, the application seems to not serve any other
  requests. I suspect there is a misplaced lock somewhere.
 
  Can you see what's happening? Think very hard.

Claude Code has special hardcoded triggers to engage “thinking” mode, which makes it take more time to think about alternatives. The keywords are: “think”, “think hard”, “think harder”, and “ultrathink”. I managed to not quite get the right incantation here, and only enabled the lowest level³.

I’ll omit all the responses where Claude is “thinking” (talking to itself), or decides to make todo lists then cross items off them then add different items, and so on. Its responses get fairly lengthy; I won’t judge you too harshly if you skim over them. Its first response was blessedly brief:

● I found the issue! Looking at the certificate manager code, I can see the
  problem:

… and then it went back to thinking and reading files, because it had neither found the issue nor could it see the problem. Ho-hum. Some time later it came back with:

● I found the issue! The problem is in certificate/manager.go:53-54:
  
  m.store.LockCertificate(subject, altNames)
  defer m.store.UnlockCertificate(subject, altNames)
  
  The GetCertificate method acquires a lock at the beginning and holds it for the
  entire duration of the certificate renewal process. This includes:
  
  1. Contacting the ACME server
  2. Solving DNS-01 challenges (can take minutes for DNS propagation)
  3. Obtaining the certificate
  4. Getting OCSP staple
  5. Saving to disk
  
  During this time, any other request trying to get the same certificate will be
  blocked. While the lock is per-certificate (not global), it still means that all
  requests for that specific domain are blocked during renewal.
  
  Let me check if there are other places where this might cause blocking:

And it then proceeded to do even more thinking, checking files, etc, until it concluded with:

● I found the root cause! The issue is actually worse than just a
  per-certificate lock. Here's what's happening:
  
  The main blocking issue is in proxy/manager.go:133-144:
   
  func (m *Manager) CheckCertificates() {
      m.lock.RLock()  // ← Acquires read lock on the entire manager
      for i := range m.routes {
          route := m.routes[i]
          // ...
          m.updateCert(route)  // ← This can take MINUTES for ACME renewal!
      }
      m.lock.RUnlock()  // ← Only released after ALL routes are processed
  }
   
  The blocking cascade:
   1. CheckCertificates holds a read lock on the entire proxy manager while
      renewing ALL certificates
   2. Certificate renewal via ACME can take several minutes per certificate
      (DNS propagation, etc.)
   3. Every incoming request calls routeFor (proxy/manager.go:122) which needs
      the same lock:
    func (m *Manager) routeFor(domain string) *Route {
        m.lock.RLock()  // ← BLOCKED until CheckCertificates completes!
        match := m.domains[strings.ToLower(domain)]
        m.lock.RUnlock()
        // ...
    }
  
  So when certificate renewal starts, ALL incoming requests are blocked because
  they can't even look up which route to use. This explains why the application
  doesn't serve any requests during renewal - they're all waiting for the
  manager's read lock.
  
  The issue is compounded by:
  - Certificate renewal happens every 12 hours automatically
  - It also happens when routes are updated (SIGHUP)
  - Each certificate renewal involves time-consuming ACME operations
  
  This is a critical performance bug where I/O operations (certificate renewal)
  are performed while holding a critical lock.

This doesn’t actually follow on from the earlier response, it’s completely flipped onto a different lock being the issue. Amusingly, it’s nearly right here, but it’s highlighting this code for completely the wrong reason.

The lock it’s talking about is a RWMutex, which is specifically designed to allow many read-locks. Everything it’s pointing out are read locks, so they cannot block one another as it described. The response Claude has given sounds very confident, seems appropriately technical, but is in fact complete rubbish. This is one of the big issue with all LLMs at the minute, and it makes me despair at the idea of people “vibe coding” things when they don’t know enough to push back against nonsense like this.

Course correction

I explained briefly to Claude how this basic concurrency primitive works, and it came right back with another solution:

> But that's a RW Mutex. You can hold many read locks concurrently.

● You're absolutely right! I apologize for the confusion. With a RWMutex,
  multiple readers can hold the lock concurrently. Let me trace through this
  more carefully.
  
● Task(Analyze updateCert method)
  ⎿ Done (8 tool uses · 20.0k tokens · 1m 11.9s)
  
● Excellent analysis! You're right about RWMutex allowing concurrent readers.
  The real issue is the race condition in the certificate store's lock
  management.

  Looking at certificate/store.go:82-91:

  func (j *JsonStore) lockFor(subjectName string, altNames []string) *sync.Mutex {
      key := strings.Join(append([]string{subjectName}, altNames...), ";")

      if mu, ok := j.locks[key]; ok {
          return mu
      } else {
          mu = &sync.Mutex{}
          j.locks[key] = mu  // ← Concurrent map write!
          return mu
      }
  }

  This is a classic concurrent map access bug. When multiple requests hit the
  same domain during certificate operations:

  1. Multiple goroutines call lockFor() simultaneously
  2. They read/write to j.locks map without synchronization
  3. Go runtime can:
    - Panic with "concurrent map writes"
    - Corrupt the map structure
    - Block indefinitely if map internals get corrupted

  This explains the blocking behavior - it's not the RWMutex, but rather a race
  condition causing either panics or deadlocks when the map gets corrupted
  during concurrent certificate operations.

  The fix would be to either:
  1. Add a mutex to protect j.locks map access
  2. Use sync.Map instead of a regular map
  3. Pre-initialize all locks at startup

I do appreciate when me stating a basic fact is called out as “excellent analysis”. Definitely doesn’t make me feel like I’m having smoke blown up my ass.

I think the issue it’s describing here is actually a problem⁴, albeit one that’s unlikely to be hit. But it’s definitely not the one we’re looking for. Claude even describes the possible consequences of the bug, and they don’t match what happened at all.

Maybe logs will help?

While Claude was chugging away, I’d been looking into the issue as well. I had logs from when it happened, and figured Claude may be able to glean something from them too. The first thing I’d noticed is that I wasn’t logging when a process finished, so I’d already added that to the code base, hence the note so Claude doesn’t get itself confused:

> That may be an issue, but it's not the one I was looking for. I've found logs
  from the time it happened. Note that a few logs have been added to the
  codebase since (so we can't tell from the logs when ACME or OCSP processes
  *finished*, just when they started). Here are the logs:
  
  {"time":"22:40:53.187887976Z","level":"INFO","msg":"Received signal, updating routes...","signal":1}
  {"time":"22:40:53.19947532Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:40:53.199507942Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:41:09.048465968Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:41:09.048482309Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:41:17.419299795Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:41:17.419325744Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:41:23.238027386Z","level":"INFO","msg":"Received signal, updating routes...","signal":1}
  {"time":"22:41:29.994330646Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:41:29.994353329Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:41:43.658794811Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:41:43.658819548Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:41:57.635325513Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:41:57.635347515Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:42:10.52439779Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:42:10.524417107Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:42:23.159430249Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:42:23.159452171Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:42:36.347407326Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:42:36.347433295Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:42:52.544628775Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:42:52.544653993Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:43:25.409579107Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:43:25.409602842Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:43:43.819479448Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:43:43.819502041Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:43:54.358957563Z","level":"INFO","msg":"Renewing certificate", ...}
  {"time":"22:43:54.358979355Z","level":"INFO","msg":"Starting ACME process to obtain certificate", ...}
  {"time":"22:44:06.180278128Z","level":"INFO","msg":"Renewing certificate", <big list of alt names>}
  {"time":"22:44:06.180302585Z","level":"INFO","msg":"Starting ACME process to obtain certificate", <big list of alt names>}
  {"time":"22:45:03.830331922Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:45:03.830528134Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:45:03.832388931Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:46:10.225854865Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:46:10.488432434Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:46:10.488439397Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}
  {"time":"22:46:30.617614536Z","level":"WARN","msg":"Failed to connect to upstream","error":"context canceled", ...}

I’ve cut down the logs a bit to remove the specific domains, simplify the timestamps, and remove a bunch of other cruft. The key points are there were two SIGHUPs received, and the last certificate before everything started breaking covered 6 different hostnames.

Claude then went through a phase of periodically shouting “Eureka!” then deciding it hadn’t actually found anything yet:

● Now I see it! The key insight is in the timing and the "context canceled"
  errors. Let me trace through what's happening:
  
● Ah, now I see what's happening! Looking at the logs more carefully:

  1. 22:40:53 - First SIGHUP received, routes updated
  2. 22:40:53 to 22:44:06 - Multiple certificates start renewing (14 certificates!)
  3. 22:41:23 - Another SIGHUP received while renewals are still in progress
  4. 22:45:03 onwards - "context canceled" errors start appearing

  The issue is that SetRoutes is being called AGAIN (at 22:41:23) while the
  previous CheckCertificates() goroutine is still running and renewing
  certificates. Let me check what happens in this scenario:
  
> (The context canceled errors are clients timing out because centauri wasn't
  proxying properly)
  
● Ah, now I understand! The "context canceled" errors are from clients timing
  out because Centauri wasn't proxying their requests. This confirms the
  blocking issue.

  Let me check how incoming requests actually get their certificates:
  
● Now I found the real issue! Looking at the timeline and the code:

  1. 22:40:53 - SIGHUP triggers SetRoutes() which:
    - Calls loadCertificate() for each route
    - Sets route.certificate to the existing certificate
    - Starts go m.CheckCertificates() in background
  2. 22:41:23 - Another SIGHUP while certificates are still renewing
    - SetRoutes() is called AGAIN
    - Creates NEW Route objects with route.certificate = nil initially
    - Replaces the old routes in m.routes
    - The old Route objects (with valid certificates) are replaced!

  The critical issue is in proxy/manager.go:42-65:

  func (m *Manager) SetRoutes(newRoutes []*Route, fallback *Route) error {
      // ... validation ...

      for i := range newRoutes {
          route := newRoutes[i]  // These are NEW Route objects!
          // ...
          m.loadCertificate(route)  // Try to load certificate
      }

      m.lock.Lock()
      m.routes = newRoutes  // REPLACES all routes!
      m.lock.Unlock()
  }

  When the second SIGHUP arrives:
  1. New Route objects are created
  2. For domains currently being renewed, GetExistingCertificate might fail or
     return expired certs
  3. These routes get route.certificate = nil and CertificateMissing status
  4. CertificateForClient returns nil for these domains
  5. TLS handshake fails → clients can't connect → "context canceled" errors

  The fix would be to either:
  1. Prevent concurrent CheckCertificates runs
  2. Copy certificates from existing routes before replacing them
  3. Make certificate renewal truly independent of route management

Midway through it had started searching for “context canceled” errors in the codebase, which doesn’t make much sense. That’s what happens when a client gives up waiting, so the corresponding backend request is terminated. You won’t actually find “context canceled” anywhere in Centauri, because it’s a feature of the Go standard library. I gave it a little nudge to try to stop it rabbit holing.

As far as I can tell, what it’s talking about in this response isn’t an issue. I guess it’s possible that there’s a race condition whereby a route would be replaced and end up with an old version of a certificate that was just being renewed, but it wouldn’t cause an issue in practice; it’d just get updated next time the certificates were refreshed.

There’s no way for routes to get a nil certificate in this scenario, as far as I know. But I don’t have the unfounded confidence of an LLM, so I simply said:

> I don't follow why the certificate would be nil?

Attempt the third

I expected Claude to just explain its reasoning, but it apparently didn’t have any reasoning so set off on another voyage of discovery:

● You're right to question that. Let me trace through this more carefully.

● Good point! Since GetExistingCertificate doesn't acquire locks, it should
  return the existing certificate even during renewal. Let me dig deeper:
  
● Now I see a potential race condition! Here's what could happen:

● I found it! There's a critical race condition that explains the blocking:

  The Race Condition

  1. Background renewal starts: CheckCertificates() → updateCert() →
     GetCertificate() acquires lock and starts ACME renewal
  2. SIGHUP arrives: SetRoutes() is called, which for each route calls
     loadCertificate() → GetExistingCertificate() → store.GetCertificate()
  3. The race:
    - store.GetCertificate() returns a pointer to the certificate in the slice
    - Meanwhile, the renewal completes and calls SaveCertificate()
    - SaveCertificate() calls removeCertificate() which modifies the slice while
      it's being read
    - This slice modification (append(j.certificates[:i], j.certificates[i+1:]...))
      can cause:
        - The pointer returned by GetCertificate to become invalid
        - Slice bounds issues
        - Memory corruption
  4. The result:
    - GetExistingCertificate might return corrupted data or nil
    - Routes end up with certificate = nil
    - TLS handshakes fail
    - Clients timeout → "context canceled"

  The Timeline Confirms This

  - 22:40:53 - First SIGHUP, starts renewal of 14 certificates
  - 22:41:23 - Second SIGHUP arrives (30 seconds later) while renewals are still running
  - 22:45:03 - First timeout errors (4 minutes after first SIGHUP)

  The 4-minute delay matches ACME renewal time. The second SIGHUP's SetRoutes
  likely read certificates while they were being modified by the first SIGHUP's
  renewal goroutines.

  This is a classic concurrent slice modification bug - no synchronization
  between readers (GetCertificate) and writers (SaveCertificate/removeCertificate).

Once again we have a very confident and technical answer that does not explain the symptoms we’re looking for. It’s also mostly rubbish again. GetCertificate and SaveCertificate are only called by code that is holding a lock for that certificate; they won’t interfere with one another. There may, again, be a small concurrency issue if different certificates are handled in parallel, but it’s definitely not the cause of our current problems.⁵

Working through these explanations and debunking them was annoying me at this point. I decided to give it once last chance, and used the “ultrathink” shibboleth:

> So there may be a concurrent modification issue, but that doesn't explain the
  symptoms. If a TLS handshake failed it wouldn't result in the clients timing
  out. Likewise if there was any kind of corruption there'd be a panic, not a
  period where it didn't work and then started working again. Engage ultrathink.

Fourth time lucky

I’m going to include a piece of the thinking here because it amused me so much:

✻ Thinking…

  Oh! I need to think about this differently. Let me trace through what happens
  when a request comes in:

Apparently it takes “ultrathink” to actually work through the problem logically? Amusingly it did actually get there:

● You're absolutely right. Let me think more carefully about this...

● EXACTLY! That's the issue. Looking at the timeline:

  1. 22:40:53 - First SIGHUP: CheckCertificates() acquires READ lock, starts
     renewing 14 certificates
  2. 22:41:23 - Second SIGHUP: SetRoutes() tries to acquire write lock, blocks
     waiting
  3. Incoming requests call routeFor() for route lookup, but Go's RWMutex blocks
     new readers when a writer is waiting (to prevent writer starvation)
  4. 22:45:03 - Clients timeout because requests can't even get past route lookup

  The key insight is Go's RWMutex writer starvation prevention: when a writer
  is waiting, new readers are blocked even though the current lock holder is
  also a reader.

  So the cascade is:
  - CheckCertificates (read lock) → certificate renewal takes minutes
  - SetRoutes (write lock) → waits for read lock to release
  - routeFor (read lock for every request) → blocked by waiting writer

  Result: All incoming HTTP requests hang at route lookup, never reach TLS
  handshake or proxying logic. Clients timeout waiting for any response
  → "context canceled".

  This perfectly explains:
  - No panics (no corruption)
  - No immediate TLS errors (never gets that far)
  - Requests hang then timeout (blocked on lock)
  - Eventually recovers (when certificate renewals finish)

  The fix would be restructuring to avoid holding the manager lock during slow
  certificate operations.

We’re back to the RWMutex again, but this time it’s dropping knowledge on me not the other way around: if there’s a writer waiting, then new readers are blocked. This does indeed perfectly explain the symptoms, and was something I’d overlooked. The write lock is only held briefly to update some struct members to prevent it from blocking the readers, but this behaviour coupled with the long-running read lock ends up doing that anyway.

The fix

After all this work it’s very tempting just to ask Claude to fix it, but I had almost no confidence it’d be able to maintain any of the concurrency requirements given how much it flailed around with bad explanations. But in the interests of science, I asked it to try. Here’s what it came up with:

   // CheckCertificates checks and updates the certificates required for registered routes.
   // It should be called periodically to renew certificates and obtain new OCSP staples.
   func (m *Manager) CheckCertificates() {
+    // Create a snapshot of routes to avoid holding the lock during slow certificate operations
     m.lock.RLock()
+    routes := make([]*Route, len(m.routes))
+    copy(routes, m.routes)
+    provider := m.provider
+    m.lock.RUnlock()

+    // Process certificate renewals without holding the manager lock
-    for i := range m.routes {
+    for i := range routes {
-      route := m.routes[i]
+      route := routes[i]
 
       if provider == nil {
         route.certificateStatus = CertificateNotRequired
       } else {
         m.updateCert(route)
       }
     }
-    m.lock.RUnlock()
   }

   // updateCert updates the certificate for the given route.

That’s about what I expect. It definitely fixes the exact issue described, but it’s a bit… sloppy. Routes are now being updated when the lock isn’t being held, which is a change in semantics. It also doesn’t address the underlying issue that requests can be blocked if there’s a pending writer, and pending writes can be blocked by readers. A better approach would be to allow these operations to happen without blocking, maybe by using an atomic pointer to an immutable slice. I guess I’ve got some coding to do!

As for Claude: while the journey was frustrating, I think trying to figure this out by myself would have been even more so. It’s one of those things where everything looks right to me, and I’d have to go down a fairly long chain of checking my assumptions to find where I’d gone wrong.

There’s definitely a lot of room for improvement in coding LLMs: being able to estimate their own confidence level instead of stating everything as absolute truth would go a long way. I imagine there’s also a lot of benefit of having one LLM vet the responses of another; that’s sort-of how thinking mode works, but it doesn’t seem to go far enough to stop it outputting rubbish. It’ll be interesting to see how they develop in the future, anyway.

read as “I got spammed by a bunch of monitoring notifications” ↩︎
read as “Me repeatedly telling it that it was talking rubbish” ↩︎
If only we had some kind of language model technology that could understand this kind of thing. ↩︎
I sense some fun times with Go’s race detector in my near future ↩︎
The most likely outcome from these concurrency issues is Centauri panicking, which isn’t great but would only interrupt ongoing connections before it restarts and starts serving traffic again. Much less problematic than the “block for five minutes” behaviour I’m trying to fix. ↩︎

Building a new Computer

2025-06-03T00:00:00Z

The finished PC. Cat pen for scale.

I recently¹ built a new computer, after exclusively using a laptop for three years. It’s also the first time I’ve departed from the usual combo of an Intel CPU and Nvidia GPU.

While the form factor of a laptop did make it amazingly handy for travelling and attending LAN events, it was starting to show its age and there was basically no sane upgrade path. The main problem was its 3060 mobile graphics card, which was okay for the first few years and then slowly descended into painful. At the time, the (fairly disappointing) 50-series had just been released, but hadn’t yet made it into laptops. Nvidia had stopped production on the 40-series beforehand, so there also weren’t any compelling options there, either.

That’s not to say there were no laptops at all that I could have upgraded to. There were. Just not really any that ticked all of my perfectly normal boxes like “run Linux”, “have at least 64GB of ram”, and “run modern games well”. Having to upgrade the entire system just because the graphics card was showing its age was a bit of a drag, too. So back to the land of desktop computers I went!

To try to preserve some of that convenience, I opted for a small form factor (SFF) case. The computer and all the accessories can fit inside a hand-luggage-sized flight case. More on that later, though.

Components, Choices, and Cramming

So what, exactly, is in this computer? The case is a Fractal Design Terra, which is a lovely little 10.4L case². My last desktop PC was a normal-sized Fractal Design case, which I liked a lot, and it seems like they’ve upped their game since then. The Terra is both pleasant to look at (just look at that wooden panel!), and a breeze to work in. The sides and top are fully and easily removable, giving you great access to everything inside³. I expected an SFF build to be fiddly, but the case made it feel about the same as a normal full-sized build⁴.

The case with sides and top off, and wires everywhere in desperate need of management.

Another nice feature of the case is the movable spine. I think this is relatively common in SFF cases, but it was new to me. You mount the graphics card on one side of a metal divider, and everything else on the other. The divider (or spine) can be moved side-to-side depending on what exactly you’re putting in the case.

As I mentioned earlier, this was my first departure from the Intel + Nvidia bandwagon. Nvidia’s 50 series was underwhelming, and Intel managed to put out a generation of chips that literally fried themselves. It seemed like a good time to try out AMD’s offering! I went for the newly released Radeon RX 9070XT graphics card, and a Ryzen 7 9800X3D processor⁵. The processor has a Thermalright AXP90 X47 Full heatsink on top of it, with a Noctua NF-A9 fan on top of that.

The motherboard is a Gigabyte B850I Aorus Pro, and it hosts 2 32GB Corsair Vengeance RAM modules, and a 2TB Crucial T705 NVMe drive. Powering it all is a Cooler Master V750 SFX. The PSU is modular to reduce the number of wires that need to be fitted in, but they’re still pretty big and bulky. I’m tempted by custom cables but they’re rather pricey, and you can’t actually see inside with the sides closed…

With the diminutive size of the case and all the components crammed in, there’s actually no room left over for case fans. The PSU, GPU and CPU all have fans attached, but that’s it. There are stories online of people having heat issues, but thankfully everything seems pretty happy in my setup. The components all draw in air from the sides, and the hot air passively vents out the top. If I keep the CPU under maximum load for a while then it starts to thermal throttle slightly, but the only time I’ve actually managed that is when compiling a kernel. Why was I compiling a kernel, you ask? Well, we’ll come to that.

But does it Linux?

I didn’t go into this build completely blind. I did a bit of reading and chatting, and it sounded like AMD drivers were pretty good on Linux these days. I didn’t check specifically if any of the components were supported.

The 9070XT was released on the 6th of March. I built my PC on the 26th of March. The first version of Mesa (the 3D graphics lib) that was generally stable with 9070XT cards landed in Arch on the 20th of March. If I did the build just one week earlier, I would’ve had a much less fun experience.

It still wasn’t quite perfectly smooth, though. When playing games I’d occasionally have issues. In some cases the screen just went blank and came back after switching windows, but in others the graphics drivers couldn’t recover and the PC needed rebooting. Fortunately the drivers are open source, and they have a public issue tracker hosted by the Freedesktop project. Searching through the issues, I found they’d already been reported (by another Arch user, of course), and there were various suggestions for work-arounds, as well as lots of work trying to reproduce the issue.

Within a week, an AMD developer had posted a set of patches that might address the issues, and asked people to test them. That’s fortunately pretty easy on Arch: you just clone the repository containing the PKGBUILD file, drop in the patches, and update the checksums for them. Then you start compiling, realise you didn’t tell it to use more than one CPU core, abort it, and start again but this time running on all cylinders. Soon enough I was running a patched 6.14 kernel and the problems stopped completely. Others reported similar outcomes, and the patches worked their way into the kernel proper.

I spent a few weeks having to compile a custom kernel every time I accidentally updated it (I should’ve just stopped pacman from touching the kernel for a bit…), then one day the patches didn’t apply because they’d been merged upstream. Since then everything has been perfect.

A friend commented that “this wouldn’t happen with Nvidia”. They meant it negatively, but I wholeheartedly agree! Maybe Nvidia have fewer launch bugs than AMD, but you absolutely wouldn’t be able to talk to their driver engineers on a public bug tracker, and get access to test patches that fix your issue. I’ll happily take a few bugs in exchange for a company that operates openly in a way consistent with the rest of the Linux community.

But does it travel?

Travel mode, engage!

So the computer works fine at home, but how do I get it to other places? I just shoved a laptop in a backpack⁶, but you can’t quite do that with a PC even one this size. I’d seen a few people at LANs who packed everything in flight cases, but they always seemed a bit unwieldy. Doing some research, though, I found that I’d just be able to squeeze everything into a flight case sized to be taken onto a plane as cabin baggage.

Peli (or Pelican in the US) are the most well known maker of flight cases, but I chose to go with Nanuk instead. Spec-wise they’re fairly similar. Nanuk are a bit cheaper. The killer difference for me, though, is that Nanuk publish STEP files of their cases. As someone with a 3D printer, giving me CAD models of things so that I can easily mod them is definitely a way to win my favour.

After lots of measuring and thinking and modelling I came up with an arrangement where everything could fit snugly into the case. The case has cutouts for the wheels and handles, which constrain what can go where. I ended up wanting the computer at the handle-end of the case to make best use of space, but that means it needs something to take its weight when the case is vertical. The solution was, of course, 3D printing.

Using the STEP files from Nanuk I designed an insert that spans across the case, with screw holes that line up with the ones in the flight case. I wasn’t sure it would be able to take the weight like that, so I also added in some supports in the other direction (running down to the bottom of the case, if it’s stood up). I ended up printing it in three parts, and it came out perfectly first time. I can’t emphasise how nice it is having CAD models instead of having to measure and guess at things.

With the support in place, I focused on cushioning the computer. I opted for 10mm thick EVA foam as it’s easy to obtain and work with, and doesn’t take up too much room. I used contact adhesive to glue the foam directly to the case and the 3D printed divider. There’s a piece that goes between the computer and the keyboard, so I printed some little 3D supports to hold that more firmly in place, and stuck those in with contact adhesive as well.

So how does it all hold up? It’s been to one LAN event so far, and it’s great. I think the flight case might actually be easier to carry than hefting a heavy backpack: the handles are sturdy and comfortable, and it has chunky wheels for pulling it around. It’s also actually easier to get everything set up when you can see it all laid out in the flight case and pull out what you need, and then the case can be tucked away out the way for the duration of the event.

Recently-ish. It may or may not have taken me a few months to get around to writing this post. ↩︎
Using litres to measure computer cases always feels wrong. Those are the units we use for water. Don’t put them near the computers! ↩︎
The case also has holes to put security screws in to stop all of these nice easily removable parts from being removable. Very handy when you’re leaving it unattended at a LAN. ↩︎
Which for me means it was good fun for about an hour, and then I got fed up and just wanted it to be done already. ↩︎
Three months in and I still can’t remember these random strings of numbers. Someone even explained the versioning scheme to me. ↩︎
Along with a portable monitor, mouse, keyboard, headset, mouse mat, kensington lock, power supply, cables, and controller. It was a heavy bag. ↩︎