Files
openclaw/docs/install/gcp.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

6.3 KiB

summary, doc-schema-version, read_when, title
summary doc-schema-version read_when title
Run OpenClaw Gateway 24/7 on a GCP Compute Engine VM with Docker 1
You want OpenClaw running 24/7 on GCP
You want a persistent Gateway on a Compute Engine VM
You need GCP provisioning, firewall, or SSH tunnel guidance
GCP

Run a persistent OpenClaw Gateway on a Debian Compute Engine VM. This page covers GCP provisioning, network access, and machine operations; the shared Docker VM runtime page owns container setup, persistence, custom binaries, verification, and updates.

Pricing varies by machine type and region. Use at least 6 GB RAM for a source image build. On a smaller machine, use the official pre-built image described in Docker VM runtime.

What you need

  • A GCP project with billing enabled
  • The gcloud CLI or the Cloud Console
  • SSH access from your laptop
  • Model and optional channel credentials
  • About 20 minutes

Provision the VM

Install the CLI from [cloud.google.com/sdk/docs/install](https://cloud.google.com/sdk/docs/install), then authenticate:
```bash
gcloud init
gcloud auth login
```

You can perform the same steps in the Cloud Console.
```bash gcloud projects create my-openclaw-project --name="OpenClaw Gateway" gcloud config set project my-openclaw-project gcloud services enable compute.googleapis.com ```
Enable billing in the
[Billing console](https://console.cloud.google.com/billing). Compute Engine
will not start without it.
| Type | Specs | Notes | | ------------- | ----------------------- | -------------------------------------- | | e2-standard-2 | 2 vCPU, 8 GB RAM | Recommended for source image builds | | e2-medium | 2 vCPU, 4 GB RAM | Use the official pre-built image | | e2-small | 2 vCPU, 2 GB RAM | Use the official pre-built image |
Create a Debian 12 VM:

```bash
gcloud compute instances create openclaw-gateway \
  --zone=us-central1-a \
  --machine-type=e2-standard-2 \
  --boot-disk-size=20GB \
  --image-family=debian-12 \
  --image-project=debian-cloud
```
Keep TCP 18789 closed to the public Internet. The SSH tunnel below needs only SSH access to the VM:
```bash
gcloud compute firewall-rules list \
  --format='table(name,network,direction,sourceRanges.list():label=SOURCE_RANGES,allowed[].map().firewall_rule().list():label=ALLOW)'
```

Restrict SSH source ranges to your administrative network when possible.
If you intentionally expose the Gateway through a reverse proxy or tailnet,
follow [Gateway security](/gateway/security) rather than adding a broad
`0.0.0.0/0` rule for port 18789.
```bash gcloud compute ssh openclaw-gateway --zone=us-central1-a ```
SSH key propagation can take a minute or two after VM creation. Wait and
retry if the first connection is refused.
On the VM:
```bash
sudo apt-get update
sudo apt-get install -y git curl ca-certificates
curl -fsSL https://get.docker.com | sudo sh
sudo usermod -aG docker "$USER"
exit
```

Reconnect so the group change takes effect, then verify the installation:

```bash
gcloud compute ssh openclaw-gateway --zone=us-central1-a
docker --version
docker compose version
```

Configure the Docker runtime

On the VM, follow Docker VM runtime from Before you begin through Verify and administer the Gateway. The maintained setup script uses these GCP host paths by default:

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

If a source build ends with Killed, ResourceExhausted, or exit code 137, resize the VM before retrying.

Access the Control UI

From your laptop, open an SSH tunnel and leave it running:

gcloud compute ssh openclaw-gateway --zone=us-central1-a -- -L 18789:127.0.0.1:18789

Open http://127.0.0.1:18789/. Paste the Gateway token from the VM's .env when prompted. To reprint the dashboard URL or approve a browser device, run on the VM:

cd openclaw
docker compose run --rm openclaw-cli dashboard --no-open
docker compose run --rm openclaw-cli devices list
docker compose run --rm openclaw-cli devices approve <requestId>

Troubleshooting

SSH connection refused

Wait one or two minutes for SSH key propagation, then retry. Check the VM is running and that an ingress firewall rule allows TCP 22 from your current network.

OS Login issues

gcloud compute os-login describe-profile

Ensure your account has Compute OS Login or Compute OS Admin Login permission.

Resize after an out-of-memory build

gcloud compute instances stop openclaw-gateway --zone=us-central1-a
gcloud compute instances set-machine-type openclaw-gateway \
  --zone=us-central1-a \
  --machine-type=e2-medium
gcloud compute instances start openclaw-gateway --zone=us-central1-a

Use a deployment service account

For personal setup, your user account is enough. Automation should use a dedicated service account with the narrowest role that works:

gcloud iam service-accounts create openclaw-deploy \
  --display-name="OpenClaw Deployment"

gcloud projects add-iam-policy-binding my-openclaw-project \
  --member="serviceAccount:openclaw-deploy@my-openclaw-project.iam.gserviceaccount.com" \
  --role="roles/compute.instanceAdmin.v1"

Avoid the Owner role. See Understanding roles.

Next steps