Files
openclaw/docs/install/docker-vm-runtime.md
Jesse Merhi 0e8faacd71 fix(scripts): build heap ignores its systemd memory budget and takes the full default (#123979)
* fix(scripts): size the tsdown heap from the build's own cgroup budget

The build heap probe only read the cgroup root (/sys/fs/cgroup/memory.max and
the v1 equivalent). Those files exist only when the process runs in a
namespaced container cgroup; under systemd the budget lives on the process's
own slice, and the v2 root carries no limit at all. So every systemd-managed
build found no limit, fell back to /proc/meminfo MemTotal, and took the full
12288 MB default heap regardless of its actual budget.

Observed on a 15.4 GiB host: openclaw-main-update.service ran tsdown with
NODE_OPTIONS=--max-old-space-size=12288 while its user@999.service slice was
bounded at 5 GiB, reaching 3.2 GB RSS and 6.25 GB peak before the host began
OOM-killing unrelated services.

Resolve the limit from /proc/self/cgroup and walk that chain instead, reading
memory.high alongside memory.max (memory.high throttles reclaim rather than
failing allocation, so a heap above it stalls the build instead of OOM-ing),
and take the tightest bound found. Root paths stay as the container fallback,
and an explicitly injected path list still disables detection.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): resolve the build heap budget from the v1 memory controller too

The slice walk only accepted the unified 0:: record, so a legacy or hybrid
systemd host fell back to the root probe and kept taking host memory. One
resolver now walks both hierarchies leaf-to-root, which makes the static root
list its own depth-0 case and removes it.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): read cgroup controller mounts instead of assuming their paths

v1 controllers can be co-mounted at the cgroup root, where memory.limit_in_bytes
sits under the slice with no per-controller directory, so the hardcoded
/sys/fs/cgroup/memory probe missed the budget and the build took the full
12288MB default. Mount points now come from mountinfo.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): translate cgroup records through the mount root

mountinfo field 4 is the subtree a cgroupfs mount exposes. Under a container
mount the /proc/self/cgroup record stays host-absolute, so walking it verbatim
probed paths below the visible mount and the build fell back to host memory.
Records now translate through the mount root before the walk.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): skip cgroup mounts that cannot represent this process

Falling back to the mount root for a record outside the mount's subtree sized
the build from an unrelated cgroup: an inherited namespace clamped the heap to
the 2048MB floor from a foreign 1GiB limit. Non-representable mounts are now
skipped, and the blind root probe only runs when no memory record exists.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): keep every cgroup mount view, not just the last one seen

One hierarchy can be visible through several mounts and only some expose a
subtree containing this process. Retaining only the last view dropped the
budget whenever a non-representable bind view came later, sending the build
back to host MemTotal.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): decode octal-escaped mountinfo paths before matching cgroups

ClawSweeper P2 on 7e64ad61f7: the cgroup resolver compared mountinfo's mount
root and mount point verbatim. The kernel escapes space, tab, newline, and
backslash in those two fields, so any cgroup mounted under such a path never
matched, the bounded slice was missed, and heap sizing silently fell back to
host memory.

Decode both fields before matching. The decoder lives in scripts/lib beside the
other shared script helpers rather than inline, so the scripts program has one
copy rather than a new ad hoc one.

Regression test fails pre-fix: a v2 mount at "/sys/fs/cgroup\040dir" with a
5 GiB memory.high yields --max-old-space-size=12288 (host fallback) before the
fix and 4352 after.

Follow-up, deliberately not bundled here: src/infra/sqlite-wal.ts,
src/commands/doctor-state-integrity.ts, and src/plugins/bundled-source-overlays.ts
each carry their own private copy of this same decoder. Consolidating all four
into @openclaw/normalization-core is the right end state, but it touches a
shared package plus three core modules and belongs in its own reviewable change.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): resolve cgroup-namespace-relative records to their mount

ClawSweeper P1 on d6fe49dd3f: inside a cgroup namespace /proc/self/cgroup
reports the namespace root ("0::/") while mountinfo field 4 stays the host
subtree the cgroupfs was mounted from ("/docker/<id>"). relativeCgroupPath then
found no prefix match and returned null; because a memory record had already
been seen, the root probe was skipped and the build fell back to host MemTotal.
A constrained container therefore missed its own budget entirely.

That namespace root is exactly what the mount exposes at its mount point, so it
resolves to "/" rather than failing closed.

Regression test fails pre-fix: a "0::/" record against a /docker/2f1a9c mount
root with a 5 GiB memory.max yields --max-old-space-size=12288 before the fix
and 4352 after.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): reject inherited cgroup mount views instead of guessing

ClawSweeper P1 on b4d200c5d2: the previous commit resolved a namespace-relative
record against any mount root, including the inherited views cgroup_namespaces(7)
documents, whose field-4 root reads "/..". Which cgroup such a view exposes is not
derivable from mountinfo, so probing it can size the build from an unrelated
cgroup's limit.

Reject non-canonical mount roots outright. An undecidable view now falls back to
host sizing, which is current main's behavior, rather than silently adopting the
wrong budget.

Regression test covers the "/.." inherited mount: it must yield host MemTotal
sizing, not the 5 GiB limit sitting behind that mount.

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(scripts): fail closed on namespace-root records against non-root mounts

ClawSweeper P1 on 731d3bbc8e: a "0::/" record does not prove that a mount
rooted at some other subtree exposes this process's cgroup. Resolving that pair
could cap the build heap from an unrelated cgroup's limit.

Return no mapping for it. An undecidable pair now falls back to host sizing,
which is current main's behavior, so the failure mode is a missed optimisation
rather than a wrong budget. The "/.." inherited-mount rejection stays; this
covers the broader ambiguous mapping it did not.

The namespace-relative test is repointed accordingly: an unrelated mounted
subtree must yield host sizing, not that subtree's limit.

Net production change: none (4 lines swapped).

Co-authored-by: jesse-merhi <79823012+jesse-merhi@users.noreply.github.com>

* fix(build): cap tsdown heap to the real budget and refuse hosts that cannot build

The 2048MB floor was applied on top of a discovered cgroup limit, so a small
container was handed a heap larger than it could honour. Measured in real
cgroups, that does not OOM-kill, it thrashes: a 1500MiB container sat pinned at
its ceiling for 10 minutes with oom_kill at 0, never finished the second of
eleven invocations, and starved every other process on the host.

Cap to the discovered budget, then refuse up front when that budget cannot hold
the build. The threshold is the whole-build peak, not a single pass: a full
eleven-invocation build peaks at 4730MiB, so a 5GiB slice completes while 4GiB
and 2816MiB slices are both killed partway through the third invocation.

The refusal runs before any output is cleaned, so a host that cannot rebuild
does not also lose the build it has.

* fix(build): harden tsdown heap admission

* fix(build): guard the default tsdown plan

* fix(build): preserve runtime-only Docker builds

* fix(build): admit only declaration cache misses

* fix(build): scope heap admission to real budgets

* fix(build): guard direct unified declarations

* fix(build): guard the canonical tsdown config

* fix(build): satisfy cache planning lint

* fix(gateway): release empty orphan leases

* fix(build): cap cgroup budget by host memory

* fix(build): serialize the canonical tsdown config

* test(build): freeze host memory fixtures

* fix(build): honor cgroup v1 soft limits

* fix(build): respect cgroup v1 hierarchy mode

* fix(build): admit unified runtime plans

* fix(build): admit every unified runtime path

* fix(build): collect repeated tsdown filters

* fix(build): ignore cgroup v1 soft limits

* fix(build): use explicit heap override as opt-in

* refactor(build): simplify memory admission

* fix(build): harden constrained build recovery

* fix(ci): prebuild runtime before real CLI shards

* fix(build): honor runtime-only runner environment

* fix(ci): satisfy tooling shard lint
2026-08-24 14:18:48 +10:00

7.7 KiB

summary, doc-schema-version, read_when, title
summary doc-schema-version read_when title
Shared Docker VM runtime steps for long-lived OpenClaw Gateway hosts 1
You are deploying OpenClaw on a cloud VM with Docker
You need the shared setup, binary bake, persistence, and update flow
Docker VM runtime

Use this runtime flow after provisioning a VM and installing Docker. Provider guides such as GCP and Hetzner own VM creation, firewall rules, SSH access, and the tunnel back to your laptop. This page owns the Docker setup shared by those hosts.

Before you begin

You need:

  • A Debian or Ubuntu VM with Docker Engine and Docker Compose v2
  • At least 6 GB RAM for a source image build; smaller hosts should use the official pre-built image below
  • The OpenClaw source checkout on the VM
  • Provider and model credentials for onboarding
  • An SSH-only or otherwise restricted provider firewall; do not expose the Gateway port directly to the public Internet

From the VM:

git clone https://github.com/openclaw/openclaw.git
cd openclaw
docker --version
docker compose version

Prepare persistent host state

The maintained setup script defaults state to the current VM user's home:

export OPENCLAW_CONFIG_DIR="$HOME/.openclaw"
export OPENCLAW_WORKSPACE_DIR="$HOME/.openclaw/workspace"
export OPENCLAW_AUTH_PROFILE_SECRET_DIR="$HOME/.openclaw-auth-profile-secrets"

Override those paths before setup if your VM uses a dedicated data disk. Keep all three directories in backups. The auth-profile secret directory contains the local encryption key for OAuth-backed auth profile token material, so it must persist but remain separate from OPENCLAW_CONFIG_DIR.

Run the maintained Docker setup

./scripts/docker/setup.sh

The script creates the host directories, builds openclaw:local, runs onboarding, generates a Gateway token, synchronizes .env, and starts the Gateway through the repository's docker-compose.yml. The Compose file pins container-side state to /home/node/.openclaw while using the host paths above as bind-mount sources.

To use an official prebuilt image instead of building from source:

export OPENCLAW_IMAGE="ghcr.io/openclaw/openclaw:latest"
./scripts/docker/setup.sh

For unattended setup, provider SecretRefs, extra mounts, sandbox setup, and all supported environment variables, use the full Docker guide.

`OPENCLAW_GATEWAY_BIND=lan` is the normal container setting: `loopback` would limit the Gateway to the container's own network namespace. Keep the published host port private with the cloud firewall, then reach it through the SSH tunnel from the provider guide.

Bake required binaries into the image

Installing binaries inside a running container is a trap: anything installed at runtime is lost on restart. Bake every external binary a skill needs into the image at build time.

The examples below cover three binaries only, alphabetically:

  • gog (from gogcli) for Gmail access
  • goplaces for Google Places
  • wacli for WhatsApp

These are examples, not a complete list. Docker Compose builds the repo-root Dockerfile, so extend that file rather than creating a standalone example or replacing its contents. The repository Dockerfile has required workspace-deps, build, runtime-assets, and final runtime stages. Its manifest extraction covers the packages/* and selected plugin workspaces before pnpm install --frozen-lockfile.

For Debian packages, prefer the existing build argument:

export OPENCLAW_IMAGE_APT_PACKAGES="socat"

For downloaded release binaries such as gog, goplaces, or wacli, add the download and install commands to the repo-root Dockerfile final runtime stage, after its package-install blocks and before USER node. Preserve the existing non-root uid 1000 setup, tini entrypoint, health check, and openclaw symlink.

The repository Dockerfile digest-pins its Node and Bun base images. Keep those reviewed pins instead of changing them to floating `FROM node:24-bookworm` references. For ARM-based VMs, choose `arm64` release assets for extra binaries; for reproducible builds, use versioned asset URLs and verify their checksums.

Rebuild the customized image without repeating onboarding:

OPENCLAW_SKIP_ONBOARDING=1 ./scripts/docker/setup.sh

If the build fails with Killed or exit code 137 during dependency installation or bundling, the VM is out of memory. Resize it before retrying.

Verify baked binaries:

docker compose exec openclaw-gateway which gog
docker compose exec openclaw-gateway which goplaces
docker compose exec openclaw-gateway which wacli

Verify and administer the Gateway

docker compose ps
docker compose logs --tail=100 openclaw-gateway
curl -fsS http://127.0.0.1:18789/healthz
docker compose run --rm openclaw-cli dashboard --no-open

/healthz returning a 200 response confirms that the Gateway process is listening. The image HEALTHCHECK polls the same endpoint. If the Control UI requires device approval:

docker compose run --rm openclaw-cli devices list
docker compose run --rm openclaw-cli devices approve <requestId>

What persists where

OpenClaw runs in Docker, but the container filesystem is not the source of truth. Long-lived state must survive restarts, rebuilds, and reboots.

Component Container location Persistence mechanism Notes
Gateway state/config /home/node/.openclaw/ OPENCLAW_CONFIG_DIR mount Includes openclaw.json, shared state, and installed plugin package roots
Agent workspace /home/node/.openclaw/workspace/ Workspace mount Code and agent artifacts
Channel credentials /home/node/.openclaw/credentials/ Config mount Channel credential material
Model auth profiles /home/node/.openclaw/agents/ Config mount agents/<agentId>/agent/auth-profiles.json
Auth-profile key /home/node/.config/openclaw/ Secret-directory mount Encryption key material; keep separate from the config mount
Skill state /home/node/.openclaw/skills/ Config mount Skill-level state
External binaries /usr/local/bin/ Docker image Must be baked at build time
Node and OS packages Container filesystem Docker image Rebuilt with the image; do not install at runtime
Docker container Ephemeral Restartable Safe to replace after mounted state is verified

Update OpenClaw

For a source-built image:

git pull --ff-only
OPENCLAW_SKIP_ONBOARDING=1 ./scripts/docker/setup.sh
docker compose run --rm openclaw-cli doctor --json

For a pinned or prebuilt image, update OPENCLAW_IMAGE to the intended tag or digest before rerunning the setup script. Routine image upgrades run startup-safe migrations against the mounted state; see Upgrading container images for recovery when a migration cannot complete automatically.