Monorepo for Tangled
blog: add microvm post
Signed-off-by: dawn <dawn@tangled.org>
+369
+369
blog/posts/spindle-microvm.md
+369
blog/posts/spindle-microvm.md
···
1
+
---
2
+
atroot: true
3
+
template:
4
+
slug: spindle-microvm
5
+
title: Spindle's new microVM engine
6
+
subtitle: How we built the new QEMU-based microVM engine
7
+
date: 2026-06-16
8
+
image: https://assets.tangled.network/blog/microvm.png
9
+
authors:
10
+
- name: dawn
11
+
email: dawn@tangled.org
12
+
handle: ptr.pet
13
+
---
14
+
15
+
Spindle gains a second engine: `microvm`. Each workflow gets its own little
16
+
virtual machine, a whole real environment you can do anything inside. It's an
17
+
upgrade from the Nixery engine while staying fully compatible with it, so if you
18
+
already have a working Nixery workflow, just change `nixery` to `microvm` and it
19
+
will work!
20
+
21
+
The interesting part is NixOS images: you configure the machine directly from
22
+
the workflow file. A few things you can do:
23
+
24
+
You can bring services up:
25
+
26
+
```yaml
27
+
services:
28
+
postgresql:
29
+
enable: true
30
+
ensureDatabases: ["spindle-workflow"]
31
+
ensureUsers:
32
+
- name: spindle-workflow
33
+
ensureDBOwnership: true
34
+
```
35
+
36
+
You can build Docker containers:
37
+
38
+
```yaml
39
+
virtualisation:
40
+
docker: true
41
+
steps:
42
+
- name: "do the thing!"
43
+
command: docker build ...
44
+
```
45
+
46
+
And you can use non-NixOS images too:
47
+
48
+
```yaml
49
+
image: alpine
50
+
steps:
51
+
- name: install golang
52
+
command: apk add go
53
+
```
54
+
55
+
It's quick on the second run, too, because it caches aggressively: your
56
+
dependencies, your services, and any other Nix derivation built inside the
57
+
microVM get pushed to spindle's Nix cache, so the next workflow that needs them
58
+
doesn't rebuild those. More on that [below](#the-nix-cache-both-ways).
59
+
60
+
And like everything else in tangled, the whole thing is self-hostable, so you
61
+
can run your own spindle with the microVM engine on your own hardware (see the
62
+
[self-hosting
63
+
guide](https://docs.tangled.org/spindles.html#self-hosting-guide)). If you want
64
+
fuller examples, there are [recipes in the
65
+
docs](https://docs.tangled.org/spindles.html#recipes) too.
66
+
67
+
## What's in a microVM
68
+
69
+
A microVM is just a VM with most of the boring parts removed. There's no BIOS,
70
+
no PCI bus to probe, no emulated graphics card, none of the slow legacy stuff a
71
+
normal QEMU machine drags along. You get virtio devices and not much
72
+
else, which means it boots very quickly and uses very little memory. Right now
73
+
QEMU is the only runner we support, but the engine is written so that other
74
+
runners (firecracker for example) can slot in later.
75
+
76
+
Inside the guest there's a small piece of software we call the agent. Spindle
77
+
never SSHes in or runs commands "from the outside"; instead the agent dials back
78
+
to spindle over vsock the moment it boots, says hello, and from then on every
79
+
step of your workflow is sent to it as a message. The agent runs the command as
80
+
an unprivileged user, streams stdout and stderr back, and reports the exit code.
81
+
The host side of this lives in
82
+
[`spindle`](https://tangled.org/tangled.org/core/tree/master/spindle/engines/microvm/agent.go)
83
+
and the guest side is a little Rust binary called
84
+
[`shuttle`](https://tangled.org/tangled.org/core/tree/master/shuttle).
85
+
(`shuttle` implements
86
+
[`agentproto`](https://tangled.org/tangled.org/core/tree/master/spindle/) which
87
+
is the protocol used by `spindle`. Technically speaking anyone could implement
88
+
this and, assuming side effects hold, you could have your own agent!)
89
+
90
+
91
+
92
+
## Two kinds of images
93
+
94
+
There are two "flavours" of image you can boot, and they're aimed at fairly
95
+
different people.
96
+
97
+
The first is **NixOS images**. These are the interesting ones: because the whole
98
+
guest is built with Nix, you can configure it from your workflow file directly.
99
+
Things like `dependencies`, `services`, `virtualisation` (e.g. Docker),
100
+
`registry` and `caches` are all written right there in the YAML, and the guest
101
+
agent builds and activates that config before any of your steps run. If we've
102
+
built that exact base plus config before, spindle can just hand the guest a
103
+
store path to realize (fetching from whatever cache `spindle` has configured)
104
+
instead of rebuilding it, so the second run is quick.
105
+
106
+
The second is **non-NixOS images**, which today just means Alpine, but can be
107
+
anything. You don't get the workflow-level NixOS config here (there's no NixOS
108
+
to configure), but if Nix happens to exist inside the image, like it does in our
109
+
Alpine one, it can still talk to the spindle Nix cache just fine.
110
+
111
+
## An example NixOS workflow
112
+
113
+
If you've used spindle before, this will look familiar: it's the same manifest
114
+
you already know, just with a few extra keys that the NixOS image understands.
115
+
Here's a workflow that needs Postgres to test against and Docker to build an
116
+
image:
117
+
118
+
```yaml
119
+
# .tangled/workflows/test.yaml
120
+
engine: microvm
121
+
122
+
when:
123
+
- event: ["push", "pull_request"]
124
+
branch: ["master"]
125
+
126
+
image: nixos
127
+
128
+
dependencies:
129
+
- go
130
+
- github:nixos/nixpkgs#hello
131
+
132
+
registry:
133
+
nixpkgs: github:nixos/nixpkgs/nixos-unstable
134
+
135
+
caches:
136
+
https://nix-community.cachix.org: "nix-community.cachix.org-1:mB9FSh9qf2dCimDSUo8Zy7bkq5CX+/rkCWyvRCYg3Fs="
137
+
138
+
services:
139
+
postgresql:
140
+
enable: true
141
+
ensureDatabases: ["spindle-workflow"]
142
+
ensureUsers:
143
+
- name: spindle-workflow
144
+
ensureDBOwnership: true
145
+
146
+
virtualisation:
147
+
docker: true
148
+
149
+
steps:
150
+
- name: run tests
151
+
environment:
152
+
PGHOST: /run/postgresql
153
+
command: |
154
+
docker build -t app .
155
+
psql -c "select 1"
156
+
go test ./...
157
+
```
158
+
159
+
The new keys each do one job:
160
+
161
+
- **`dependencies`** are the packages your steps get to use. They go into a
162
+
`mkShellNoCC` devshell that every step sources before it runs, so you get the
163
+
whole stdenv environment (setup hooks like `pkg-config` wiring up
164
+
`PKG_CONFIG_PATH`, etc.) and not just the bare binaries. That means you can
165
+
use a dependency like `openssl` and compile the `openssl-sys` Rust crate
166
+
without pain! A bare name like `go` is looked up in nixpkgs (same as Nixery),
167
+
but you can also point at any flake with the `flakeref#attr` syntax, so
168
+
`github:nixos/nixpkgs#hello` pulls `hello` straight out of that flake.
169
+
- **`registry`** is how you remap the global refs. Here we pin `nixpkgs` to
170
+
`nixos-unstable`, so now the bare `go` above resolves from unstable. You can
171
+
alias your own flakes the same way (`myflake: github:me/x`, then
172
+
`myflake#tool` in `dependencies`).
173
+
- **`caches`** is a map of binary cache URL to its trusted public key. They get
174
+
wired into the read proxy (more on that just below), so the guest can
175
+
substitute prebuilt paths from them instead of building everything from
176
+
scratch.
177
+
178
+
`services` and `virtualisation` are the interesting parts: they're passed
179
+
straight through to NixOS, so anything you could write in a NixOS config you can
180
+
write here. `services.postgresql.enable` brings Postgres up before any of your
181
+
steps run.
182
+
183
+
Since steps run as the `spindle-workflow` user, naming a database after that
184
+
user with `ensureDBOwnership` is the easy path to a working DB -- Postgres peer
185
+
auth maps the unix user straight to the matching role, so `psql` connects over
186
+
the socket with no password and no extra setup (this name-matching is a NixOS
187
+
requirement for `ensureDBOwnership`, if you want a differently named DB you'd
188
+
grant access yourself).
189
+
190
+
`virtualisation.docker: true` is shorthand for `virtualisation.docker.enable =
191
+
true`, which gets you a real Docker daemon inside the VM. By the time your first
192
+
step runs, Postgres is listening and the Docker socket is there, no sidecar
193
+
dance, it's just part of the machine.
194
+
195
+
(`true` works as shorthand for `.enable = true` anywhere an `enable` option
196
+
exists, so most "just turn this on" services are a one-liner!)
197
+
198
+
## The architecture
199
+
200
+
201
+
202
+
### Nix cache, both ways
203
+
204
+
Spindle talks to its Nix cache through two proxies that run on the host, so the
205
+
guest never needs credentials or direct network access to reach it. Like the
206
+
agent, they use vsock to talk to spindle.
207
+
208
+
The read proxy fans out to the configured substituters plus any caches you
209
+
listed in your workflow, so when the guest needs to realize a store path it asks
210
+
the proxy and the proxy fetches it. The request is sent concurrently to the read
211
+
caches, so the one that answers it first wins.
212
+
213
+
The upload proxy goes the other way: any path built inside the guest gets pushed
214
+
back out to spindle's Nix cache (if one is configured), so the next workflow
215
+
that needs it doesn't have to build it again. Any paths that already exist on
216
+
any of the configured read caches won't be uploaded. As the agent reports
217
+
built paths, they're queued and uploaded in the background while the rest of the
218
+
workflow keeps running, so uploads overlap with work instead of blocking it. If
219
+
any are still in flight when we reach VM teardown, the workflow waits until
220
+
everything has drained.
221
+
222
+
Spindle can be configured to use `http`, `ssh-ng` or `ssh` URLs as a binary
223
+
cache to upload to, so for example, `ssh-ng://localhost` would just upload to
224
+
the local Nix store on the machine that the spindle runs on! `ssh-ng` and `ssh`
225
+
require Nix to be present in PATH so that the spindle can use `nix copy` to
226
+
upload to them, but if you are using a binary cache that supports `http` (for
227
+
example, [ncps](https://github.com/kalbasit/ncps)) Nix does not need to be
228
+
present.
229
+
230
+
### Building the images
231
+
232
+
Image builds are done with Nix. For NixOS we lean on
233
+
[microvm.nix](https://github.com/microvm-nix/microvm.nix) and layer our own bits
234
+
on top (stripping down kernel modules, configuring users, etc.). For Alpine
235
+
there's a smallish Nix definition that fetches the kernel, the initrd and the
236
+
kernel modules, sets up an init script that configures the machine on boot,
237
+
copies in the dependencies we want (`nix`, `git`, etc.) and compresses the whole
238
+
rootfs into a squashfs.
239
+
240
+
None of this *has* to be Nix, though. As far as spindle is concerned an image is
241
+
valid as long as a few things hold: a guest agent (that implements `agentproto`)
242
+
is present and gets started on boot, a `spindle-workflow` user exists, and the
243
+
work directory is set up at `/workspace`. That can be built however you like.
244
+
245
+
### Finding an image
246
+
247
+
Every built image ships a `spec.json` next to its artifacts. The spec is the
248
+
whole contract: where the kernel and initrd and read-only store disk live, the
249
+
boot args, how much memory and how many vCPUs to give it, the shell to run steps
250
+
in, the writable volumes, the network interfaces, and the runner-specific knobs
251
+
(machine type, CPU, extra QEMU args). NixOS images also carry a `baseConfigHash`
252
+
identifying the base config baked in (this is the hash of
253
+
`nixosSystem.config.system.build.toplevel.outPath`).
254
+
255
+
A workflow picks an image with the `image` key at the top level. The name is
256
+
matched literally against what's on disk, we look for a directory called
257
+
`<name>` with a `spec.json` in it, then fall back to a flat `<name>.json`. The
258
+
nice property here is that resolution depends *only* on the name and what's on
259
+
disk, never on the host doing the resolving, so the same workflow resolves to
260
+
the same image on every spindle. If an operator keeps multiple arches side by
261
+
side they can name them `nixos-x86_64`, `alpine-aarch64` and so on (that suffix
262
+
is just part of the name, it's not handled specially). If you want, for example,
263
+
`nixos` to work, you can just symlink `nixos` to `nixos-x86_64`.
264
+
265
+
Right before launch we double-check the referenced files actually exist
266
+
and that the host has the tools we need: `mkfs.ext4` for the volumes, the
267
+
QEMU binary for the spec's arch, `/dev/kvm` and `/dev/vhost-vsock`, plus
268
+
the `ip` / `mount` / `slirp4netns` / `unshare` toolchain if the image
269
+
wants networking.
270
+
271
+
### The life of a workflow
272
+
273
+
A workflow moves through a handful of stages: it gets parsed and its
274
+
image resolved, it waits for a slot, it gets set up, its steps run, and
275
+
then everything is torn down.
276
+
277
+
The waiting bit matters a lot. Each image declares how much memory, how many
278
+
vCPUs and how much disk it needs, and a workflow has to acquire a slot from a
279
+
resource scheduler before anything boots. The scheduler is work-conserving with
280
+
aging and per-user fairness, so one person submitting a hundred jobs won't
281
+
starve everyone else, and slots don't sit idle if there's work that fits in the
282
+
budget.
283
+
284
+
Once a slot is acquired, we do the setup. Spindle allocates a random vsock CID
285
+
for the guest and registers it with the agent hub. It creates the per-workflow
286
+
work directory, starts the two cache proxies (described earlier), a DNS proxy
287
+
that resolves through the host and filters out private/special-use addresses,
288
+
then creates the VM: writable volumes become sparse files formatted ext4, the
289
+
store disk is attached read-only, and QEMU is started with `-sandbox on`,
290
+
`-nodefaults`, no display, no monitor, etc. with serial (on boot) /
291
+
`virtio_console` output to a log file and a QMP socket for control.
292
+
293
+
Then we wait for the machine. We poll QMP until QEMU says the guest is running,
294
+
then wait for the agent's handshake to arrive over vsock from the CID we expect.
295
+
The agent tells us its protocol and versions, and spindle sends back the job id,
296
+
the trusted cache public keys, and the cache and DNS proxy ports. From there
297
+
steps run one at a time as `$shell -lc <command>`, as the unprivileged workflow
298
+
user in `/workspace/repo`, with the right environment and any unlocked secrets.
299
+
If the workflow activates a NixOS config and we've already built that exact base
300
+
plus config, the activation step can realize a cached toplevel store path instead
301
+
of rebuilding. Either way, whether it's building the config fresh or pulling a
302
+
cached toplevel down, that output streams straight into the activation step's log
303
+
as it happens, so you can watch the closure come in instead of staring at a blank
304
+
screen wondering if anything's happening.
305
+
306
+
Timeouts are cooperative: we work out a deadline from the workflow timeout
307
+
and send it to the guest, with a little grace on our side so the guest
308
+
gets a chance to report the timeout itself rather than us just yanking the
309
+
machine out from under it. And if the VM crashes mid-step we tail the
310
+
serial and QEMU logs into the step's stderr, because "guest agent
311
+
connection lost: EOF" is a genuinely useless thing to read at 2am...
312
+
313
+
Teardown is the same whether the workflow passed, failed or timed out:
314
+
drain any pending Nix cache uploads, ask the agent to power off, wait for
315
+
QEMU to exit (falling back to a QMP `system_powerdown`, and finally a
316
+
kill if it's being stubborn), then close the proxies and remove the work
317
+
directory.
318
+
319
+
### Locking down the network
320
+
321
+
A VM that can reach the host's local network is a VM that can reach things it
322
+
has no business reaching. So QEMU doesn't run in the host's network namespace at
323
+
all. We `unshare` into fresh user, net and mount namespaces first. Inside that
324
+
namespace a small wrapper bind-mounts a resolv.conf pointing at `127.0.0.1` so
325
+
that QEMU's built-in slirp DNS isn't used, then installs blackhole routes for
326
+
every special-use IP range (RFC 6890, so private networks, link-local, loopback,
327
+
etc.) before it execs QEMU. `slirp4netns` then provides the namespace's outbound
328
+
internet connection, with `--disable-host-loopback`, sandbox and seccomp all on.
329
+
QEMU runs *inside* that namespace, and the guest's network card is attached to
330
+
QEMU's own built-in user-mode networking. So every packet from the guest takes
331
+
two hops: guest → QEMU's slirp → the namespace's `slirp4netns` → the internet.
332
+
The guest never sees the host's network and the host's network never sees the
333
+
guest. All of this is done without needing any privileges!
334
+
335
+
Guest DNS doesn't use either slirp layer. The guest's `/etc/resolv.conf` points
336
+
at shuttle on `127.0.0.1:53`, and shuttle forwards DNS packets over vsock to
337
+
the host-side DNS proxy. That proxy resolves through the host's real resolver
338
+
and strips any answers that point at private or special-use addresses, so guest
339
+
traffic can only ever reach the outside world, never the host or anything on its
340
+
local networks.
341
+
342
+
### Budgets and cgroups
343
+
344
+
The scheduler's budget is bookkeeping on its own, it tracks what it's handed
345
+
out, and the runner (QEMU) will ensure that a workflow only gets those. But
346
+
optionally the whole thing (QEMU and slirp4netns both) gets placed in a
347
+
per-workflow cgroup with memory, swap etc. limits, which is an extra enforcement
348
+
layer on top, considering QEMU and slirp4netns themselves also use resources. A
349
+
nice side effect is that when the cgroup OOM-kills the VM we can see that it was
350
+
an OOM and report it as such, instead of surfacing it as a generic crash and
351
+
leaving you guessing.
352
+
353
+
The spindle itself also gets a cgroup with `memory.min` set, which means that in
354
+
a host OOM situation, it should be the workflows that die first, not the spindle
355
+
itself.
356
+
357
+
## On the roadmap
358
+
359
+
A few things that are coming next:
360
+
361
+
- [firecracker](https://github.com/firecracker-microvm/firecracker) runner
362
+
support. QEMU microVMs are good and all, but firecracker VMs are more
363
+
efficient to run concurrently and are leaner overall.
364
+
- ssh-on-fail: when a workflow fails, you should be able to ssh in to debug why.
365
+
This can be really useful in situations where you need just *a little* bit
366
+
more info if something unexpected fails so you don't sit around there running
367
+
the workflow 10 times over.
368
+
369
+
Feel free to come and ask any questions you might have on https://chat.tangled.sh!