What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
If OpenClaw is not working on your VPS, start with its built-in diagnostics, then match the failure to the service, connection, channel, or file-permission layer. The fastest first check is the official command ladder below; it helps distinguish a missing command from a stopped Gateway, a disconnected dashboard, a channel issue, or a Docker ownership error. Keep Gateway port 18789 off the public internet while troubleshooting.
What should I check first when OpenClaw is not working?
Run the OpenClaw troubleshooting commands in order, beginning with the sanitized triage report. This gives you a quick view of the installation, Gateway, channels, and logs before you make configuration changes. OpenClaw’s troubleshooting runbook calls this the first-60-seconds ladder:
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →openclaw triage
openclaw status
openclaw status --all
openclaw gateway probe
openclaw gateway status
openclaw doctor
openclaw channels status --probe
openclaw logs --followopenclaw triage creates a sanitized diagnosis and can create a support archive when the Gateway is reachable. The remaining commands provide progressively more detail: overall status, a Gateway connectivity probe, service status, checks for installation or configuration problems, live channel probes, and streaming logs. Don’t assume a successful command alone means the whole setup works: a healthy result should show a running Gateway, successful connectivity, a plausible capability level, live channel transport where supported, and no repeating fatal errors in the logs.
Read the ladder as a dependency chain. If the executable itself cannot run, fix PATH before interpreting service output. If the executable works but the Gateway probe fails, stay at the service and listener layer. If the Gateway is healthy but a channel probe or model check fails, focus on credentials, pairing, allowlists, or outbound networking. Finally, use the live log stream while reproducing the problem; a timestamped error that appears exactly when you send a message or reload the dashboard is more useful than an old warning.
#1 Best Overall
- Includes Raspberry Pi 5 with 2.4Ghz 64-bit quad-core CPU (8GB RAM)
- Includes 128GB Micro SD Card pre-loaded with 64-bit Raspberry Pi OS, USB MicroSD Card Reader
- CanaKit Turbine Black Case for the Raspberry Pi 5
- CanaKit Low Noise Bearing System Fan
- Mega Heat Sink - Black Anodized
If a command fails, note its exact error and continue with the section below that matches it. For example, “command not found” is an installation-path problem, while a running Gateway with no replies points toward model authentication, channel access rules, or network reachability. Capture the command, relevant output, host operating system, and whether the process runs directly, under a service manager, or in Docker. Remove tokens and other secrets from anything you share.
Why is the openclaw command not found?
OpenClaw’s installation guide says this is almost always a PATH problem: the command may be installed, but your shell cannot find npm’s global executable directory. Check the Node.js version and npm’s global prefix, then verify that the corresponding global bin directory is included in your PATH.
- Check Node.js with
node -v. - Inspect npm’s global prefix with
npm prefix -g. - Check that npm’s global bin directory is included in your shell’s
PATH. If it is missing, update the relevant shell configuration and open a new shell so the change takes effect. - Confirm the repair by running
openclaw --version,openclaw doctor, andopenclaw gateway status.
The OpenClaw installation guide is the reference for installation and PATH checks. If node -v also fails, resolve the Node.js installation or shell environment first; OpenClaw’s executable cannot be diagnosed reliably until that prerequisite is available.
Be aware that a shell used interactively may have a different PATH from the environment used by systemd, launchd, a Scheduled Task, or a container. A command can therefore work in your terminal while the managed service reports that it cannot find OpenClaw. After correcting the interactive PATH, check the service definition’s environment as well and restart the service through its normal manager. Avoid installing a second copy merely because one shell cannot see the first; duplicate installations make version and configuration problems harder to trace.
Why won’t the Gateway service start?
Check the Gateway’s reported status and watch its logs while starting or restarting it. The common signatures in the official runbook point to three different causes: a missing or non-local gateway.mode, a refused bind because an externally reachable listener lacks authentication, or another process already occupying the port.
- Run
openclaw gateway statusto see whether the service is stopped, running, or reporting an error. - Run
openclaw logs --followand look for the first relevant startup error rather than focusing only on later repeated messages. - If the message indicates missing or non-local
gateway.mode, inspect the Gateway configuration and set its mode as required for the intended deployment. - If the Gateway refuses to bind without authentication, avoid exposing it without authentication. Prefer a loopback bind and secure remote access rather than making the listener public.
- If logs say
another gateway instance is already listeningor showEADDRINUSE, identify and resolve the existing listener instead of repeatedly starting additional Gateway instances.
The service manager depends on the host: managed Linux deployments normally use a systemd user service; macOS uses launchd; native Windows uses a Scheduled Task or startup fallback. OpenClaw’s installation index describes deployment targets, and the troubleshooting runbook covers the startup signatures. A service that starts manually but not after reboot should be investigated through the host’s service manager rather than treated as a channel or dashboard fault.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsDo not keep changing several settings between attempts. Record the original configuration, correct one likely cause, then start the Gateway once and inspect the first new log message. If a second process owns the port, stopping it may affect another application; verify what is listening before terminating anything. Conversely, repeatedly launching OpenClaw by hand can create the exact duplicate-instance condition you are trying to fix.
When the Gateway starts slowly rather than failing immediately, distinguish a timeout from a crash. A service manager may declare failure while OpenClaw is still initializing. The VPS tuning options below can extend startup time, but they cannot repair an invalid mode, missing authentication, or an occupied port.
Rank #2
- Includes Raspberry Pi 4 4GB Model B with 1.5GHz 64-bit quad-core CPU (4GB RAM)
- Includes Pre-Loaded 32GB EVO+ Micro SD Card (Class 10), USB MicroSD Card Reader
- CanaKit Premium High-Gloss Raspberry Pi 4 Case with Integrated Fan Mount, CanaKit Low Noise Bearing System Fan
- CanaKit 3.5A USB-C Raspberry Pi 4 Power Supply (US Plug) with Noise Filter, Set of Heat Sinks, Display Cable - 6 foot (Supports up to 4K60p)
- CanaKit USB-C PiSwitch (On/Off Power Switch for Raspberry Pi 4)
Why does the dashboard keep disconnecting?
First verify that the dashboard is reaching the Gateway endpoint you intended, with the right port and authentication token. Then check whether the route used to reach that endpoint—such as an SSH tunnel or Tailscale connection—is still active. The browser’s connection errors can identify a wrong URL, rejected origin, insecure context, or stale credentials.
- Run
openclaw gateway statusandopenclaw statusto confirm the Gateway is up. - Check the dashboard URL and port against the Gateway endpoint. If remote access depends on an SSH tunnel or Tailscale, verify that connection is active.
- Confirm that the dashboard has the correct token. If the browser was paired earlier, stale pairing or credentials may also need attention.
- Run
openclaw logs --followwhile reconnecting, then map the log message to the likely cause below.
| Log message | What it points to | What to check |
|---|---|---|
device identity required |
An insecure HTTP context | Whether the dashboard is being opened in a context that supports the required device identity. |
origin not allowed |
The browser Origin is not allowed | The dashboard origin and the Gateway’s allowed-origin configuration. |
AUTH_TOKEN_MISMATCH |
A token-mismatch path | Whether the dashboard and Gateway are using matching credentials. |
Repeated unauthorized |
Wrong credentials or stale pairing | The supplied credentials and the current pairing state. |
gateway connect failed |
The UI is targeting the wrong endpoint or an unreachable Gateway | The URL, port, Gateway status, and remote route. |
These signatures and checks are described in OpenClaw’s troubleshooting runbook and logging and debugging FAQ. Change one relevant connection or authentication setting at a time so you can tell which correction resolves the disconnect.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →A disconnect that occurs only after a period of inactivity can still be a route problem: confirm that the SSH tunnel or Tailscale session remains established, then watch the Gateway log while refreshing. A disconnect that affects every browser and device is more likely to be the Gateway, token, origin, or route; one device alone may have stale pairing or cached credentials. Keep the dashboard’s endpoint and token consistent across devices instead of generating ad hoc public URLs.
Why is OpenClaw running but not replying?
A running Gateway does not prove that a model can authenticate, that the sender is permitted to reach a channel, or that the channel transport is connected. Check model status, probe channel connectivity, and inspect logs together; they help separate those failure points.
- Run
openclaw models statusto check model configuration and authentication on the Gateway host. - Run
openclaw channels status --probeto test channel transport where supported. - Run
openclaw logs --followwhile sending a new message, then look for authentication failures, blocked senders, or channel errors. - Check whether the sender is allowed by the channel’s pairing or allowlist rules.
- If the setup relies on a remote tunnel or Tailscale path, verify that route is available to the Gateway host.
Model credentials must be available on the Gateway host, not merely on a separate computer where you administer OpenClaw. Pairing and allowlist rules can also block an otherwise healthy sender. OpenClaw’s logging and debugging FAQ identifies these as common reasons the service runs without delivering replies.
Use a controlled test message after each change. If the model check fails, fix that before changing channel policy; if the model is authenticated but the channel probe fails, investigate the channel route and transport; if both pass but a sender is silent, inspect pairing and allowlist decisions in the logs. This order prevents a channel rule from masking a model-credential problem or vice versa.
What if Telegram command registration fails?
For a Telegram setMyCommands error, check channel logs rather than treating it as a Gateway-startup problem. The error BOT_COMMANDS_TOO_MUCH can require reducing the number of commands; a fetch failure instead calls for checking outbound HTTPS and DNS access to api.telegram.org. These diagnoses are in the same OpenClaw logging and debugging FAQ.
That distinction matters on locked-down VPS networks. A service may accept local requests while outbound DNS or HTTPS is blocked, so test the host’s egress path and resolver as well as OpenClaw’s channel status. Do not expose the Gateway publicly just to make a channel reachable.
Rank #3
- Not including the Raspberry Pi 5 (8GB), the Crowpi advanced version comes with the Raspberry Pi 5
- ELECROW Black Case for the Raspberry Pi 5, CrowPi is equipped with a 9-inch HD touchscreen along with a camera; All the regular components used in DIY electronics are packed into the CrowPi development board, such as LCD, LED matrix, buzzer, light sensor, PIR sensor, ultrasonic sensor, IR sensor, etc
- Raspberry Pi Sensors: The Crowpi raspberry pi 5 programming kit is jam-packed with lots of buttons such as 19 different sensors in a tidy easy to use package; You don't have to wait and wire things
- Build Quality: Solid ABS shell and well made components in one place make it strong and convenient to travel
- Programming Lessons: This raspberry pi 5 learning kit ships with step by step instructions and provides 21 lessons to take you through identifying components reading code and running it in the terminal
How do I fix Docker plugin ownership or EACCES errors?
A plugin may be blocked when the files it needs are owned by a different Unix user from the process running OpenClaw. For the Docker installation described in the runbook, the process runs as UID 1000, so the configuration and workspace bind mounts should be owned by that UID and group.
For a UID 1000 Docker install, repair the bind-mounted paths and then run the doctor:
sudo chown -R 1000:1000 /path/to/openclaw-config /path/to/openclaw-workspace
openclaw doctor --fixReplace the example paths with the actual host paths mounted into the container. The recursive ownership change affects all files below those paths, so check that they are the intended OpenClaw configuration and workspace directories before running it.
If you intentionally run OpenClaw as root, the expected owner differs. The runbook’s root-specific repair is:
sudo chown -R root:root /path/to/openclaw-config/npm
openclaw doctorUse the repair for the identity that actually runs the process; applying the UID 1000 fix to an intentional root deployment, or the root fix to the normal Docker process, can leave ownership mismatched. See the official troubleshooting runbook for the ownership guidance.
Ownership is only one Docker layer. Also verify that the host paths are the same paths mounted into the container, that the container is running the expected user, and that the plugin directory is not being replaced by a read-only mount. Make a backup before a recursive change, especially when a workspace contains files created by other services. If the error returns after a container rebuild, inspect the image and volume user settings rather than repeatedly changing ownership by hand.
Free tools Windows power users keep installed
One-click scans. No signup required.
How can I make a VPS start and restart more reliably?
OpenClaw’s VPS guide gives several host-tuning options for startup tracking and recovery. They are options to consider for a particular host—not universal minimum requirements—and should be applied only when they address a symptom you have observed.
OPENCLAW_NO_RESPAWN=1can simplify process tracking.- Setting
NODE_COMPILE_CACHEto a writable cache path can improve repeated startup. - On a slow host, a longer systemd
TimeoutStartSec, such as 90 seconds, can allow more startup time. - A managed policy of
Restart=alwayswithRestartSec=5can restart the service after it exits. - SSD-backed state and cache paths can reduce random-I/O cold-start penalties.
These settings address different conditions: process-tracking simplicity, repeated compilation work, slow service startup, unexpected exits, and storage latency. Use the host’s service configuration and OpenClaw logs to identify which applies. The exact suggestions are in the OpenClaw VPS deployment guide; it does not establish them as required settings for every VPS.
Rank #4
- Fully assembled for plug-and-play operation
- Includes Raspberry Pi 5 with 8GB RAM
- 256 GB PCIe Pi NVMe SSD (Pre-loaded with Pi 64-Bit OS)
- M.2 HAT+
- CanaKit Turbine Black Case for the Pi 5
A restart policy is recovery, not diagnosis. If the Gateway exits because authentication is missing, the port is occupied, or configuration is invalid, Restart=always can produce a restart loop and noisy logs. First obtain one clean failure, correct the underlying cause, and then use a managed restart policy for unexpected exits. Keep state and cache on reliable, writable storage; SSD-backed paths are particularly helpful when startup performs many small reads and writes.
When evaluating a VPS, compare the provider’s CPU and RAM headroom, storage I/O, network reachability, backup options, and maintenance controls rather than selecting only by advertised price. The OpenClaw guidance supplies no universal CPU, RAM, disk, or monthly-cost minimum, because workload, channels, model use, and retention change the requirement.
How should I secure OpenClaw on a VPS?
Keep Gateway port 18789 off the public internet and bind the Gateway to 127.0.0.1 where that fits the deployment. For remote access, route through Nginx, an SSH tunnel, or Tailscale rather than exposing the Gateway directly. The hosting guidance also recommends backing up the workspace and configuration files.
- Check the listener address and firewall so port 18789 is not publicly reachable.
- Choose a controlled remote-access route—Nginx, SSH tunneling, or Tailscale—and verify it is active when dashboard access is needed.
- Back up both workspace and configuration files so recovery includes working data as well as setup.
- Treat an exposed Gateway and broad file or shell access as serious security risks.
The hosting checklist is described in the VPS hosting guide. A bind address, authentication, and remote-access path solve related but distinct problems: a private listener limits network exposure, while a secure route provides a deliberate way to administer it remotely.
Test remote access from the same kind of network your users will use, but keep the private bind in place. If you use Nginx, SSH, or Tailscale, document the endpoint and recovery steps so a reboot does not leave you guessing whether the Gateway or the tunnel is down. Back up before changing authentication or origin settings, and never paste live tokens into tickets or public logs.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallShould I host OpenClaw on a VPS, a local computer, a Raspberry Pi, or in Docker?
There is no single best host in the available deployment guidance. OpenClaw lists cloud servers, VPS, Docker, Kubernetes, Raspberry Pi, and other targets; decide by weighing always-on reliability, CPU and RAM headroom, storage I/O, network reachability, maintenance, remote-access security, and backup recovery.
| Option | What to evaluate |
|---|---|
| Local computer | Whether it stays on when replies are needed, its available CPU/RAM and storage performance, and how remote access will be secured. |
| Raspberry Pi | Whether the selected device and storage provide enough headroom for the intended workload, and how you will keep it reachable, maintained, and backed up. |
| Docker | Container and host networking, bind-mount paths, and ownership matching the process UID; ownership mismatch can cause plugin or EACCES failures. |
| VPS or cloud server | Always-on availability, resource and storage capacity, service restart behavior, secure remote access, and backups. |
The installation guide lists supported deployment categories, while the VPS guide covers VPS-specific tuning. The available guidance does not provide a universal minimum CPU, RAM, storage size, or cost for these options, so do not infer that a particular device or server tier is sufficient without checking the requirements for your intended workload.
Local hardware can be economical when it is already powered on, but uptime, inbound reachability, and backup responsibility remain yours. A Raspberry Pi can provide a compact always-on host, yet storage quality, cooling, power stability, and remote administration deserve the same scrutiny as a VPS. Docker improves repeatability and isolation, but adds container networking, volume, and UID decisions. A VPS generally simplifies always-on reachability and service management, while adding a recurring provider charge and an obligation to secure the public-facing host. Compare total maintenance and recovery effort, not just purchase or rental price.
Best Value
- 【What you Get】You will get 1*Pi 5 8GB Single Board,1*RasTech Case,1*Active Cooler,1*Screwdriver,1*Installation instructions,12-month free warranty, lifetime service, 24-hour prompt and friendly response.
- 【More Connectors】There are two USB 3.0 ports(5Gbps simultaneously) and two USB 2.0 ports, which triple total bandwidth ,support any combination of up to two cameras or displays. Peak SD card performance is doubled through support for the SDR104 high-speed mode. It provides a smooth desktop experience for you. Offer Gigabit Ethernet and a PCIe interface, along with dual-band Wi-Fi and Bluetooth 5.0/BLE wireless capability. The RasTech Pi 5 Kit use the new 27W 5.1V 5A USB-C power connector.
- 【 Support Dual 4Kp60 Display 】Each of the two microHDMI sockets can control a 4K display at 60 Hertz, now support HDR, offering super HD video for media streaming projects. RPi 5 is the first RPi model that comes with a PCI Express port (PCIe 2.0 x1 with 500 MB/s) to attach SSDs (requires separate M.2 HAT).
- 【 Excellent Chips And Applications】Pi 5 is a full-size Pi computer using silicon built in-house at Pi. The RP1 “southbridge” provides the bulk of the I/O capabilities for Pi 5. Pi 5 is more friendly and convenient in the development of Internet of Things, Web development, machine identification, automatic control and other electronic equipment applications and network.
- 【 Faster CPU, Better GPU 】 Pi 5 features a Broadcom BCM2712 64-bit quad-core Arm Cortex-A76 processor running at 2.4GHz, it delivers a 2–3× increase in CPU performance relative to RaspberryPi 4. The 800MHz VideoCore VII GPU is compatible to OpenGL ES 3.1 and Vulkan 1.2, substantial uplift in graphics performance. Pi 5 Offers lightning-fast CPU speed, a PCI Express interface, a Real Time Clock (RTC) and a power button and runs significantly cooler than Pi 4.
What is a practical recovery sequence after a host change?
Before migrating or rebuilding, back up the OpenClaw workspace and configuration files. On the new host, verify Node.js and PATH, restore the files, confirm ownership for the selected process identity, and run the first-minute command ladder. Keep the Gateway private while you validate status, model authentication, channel probes, and dashboard access through a controlled route.
- Record the current deployment type, service manager, listener address, port, remote-access method, and process user.
- Back up workspace and configuration data, excluding secrets from any copy shared for support.
- Install the required runtime and confirm
node -v, npm’s global prefix, andopenclaw --version. - Restore paths with ownership matching the service or container user; use the documented UID 1000 or root repair only when it matches your deployment.
- Run
openclaw doctor, then check Gateway status and logs before enabling remote access. - Probe models and channels, send a controlled test message, and only then make the dashboard reachable through Nginx, SSH, or Tailscale.
This sequence limits variables. It also gives you a clean before-and-after record if a provider, operating-system, storage, or networking change introduces a new failure.
Frequently Asked Questions
What should I include in an OpenClaw support report?
Start with the sanitized output from openclaw triage; it can create a support archive when the Gateway is reachable. Include the relevant error signature and the diagnostic command that produced it, plus the host and deployment type, while removing credentials, tokens, and private data.
Does OpenClaw need to be publicly reachable for remote dashboard access?
No. The hosting guidance recommends keeping port 18789 off the public internet and using a controlled route such as an SSH tunnel, Nginx, or Tailscale for remote access.
Can a VPS restart policy fix a Gateway that never starts?
A restart policy can bring a process back after it exits, but it does not correct a configuration, authentication, or occupied-port error that prevents startup. Inspect Gateway status and logs first to identify the failure.
Recommended Free Tools
What files should I back up before changing an OpenClaw host?
The VPS hosting guidance specifically recommends backing up the workspace and configuration files. Those preserve both your work and the setup needed to restore the service.
Why does OpenClaw work in my shell but fail as a service?
The interactive shell and the service manager can have different PATH and environment settings. Verify npm’s global bin directory in the service environment, then check the service’s process user and restart it through systemd, launchd, a Scheduled Task, or the appropriate host manager.
How do I tell a model problem from a channel problem?
Run openclaw models status and openclaw channels status --probe while following the logs. Missing model authentication points to the Gateway host; a failed channel probe, pairing decision, allowlist rule, or unavailable tunnel points to channel access or transport.
What does EADDRINUSE mean for OpenClaw?
It means the configured listener port is already occupied. Check which process owns it, determine whether another Gateway instance is running, and resolve the existing listener before starting a second instance.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsIs a specific VPS size or monthly price required?
No universal CPU, RAM, storage, or cost minimum is given. Choose capacity and spending based on workload, always-on needs, storage I/O, network reachability, backups, and the maintenance you can provide.
The Bottom Line
Start with the first-minute command ladder, then isolate the layer that fails: PATH, Gateway startup, dashboard route and credentials, model or channel access, Docker ownership, or host reliability. Keep port 18789 private, use a controlled remote-access path, back up workspace and configuration files, and change one variable at a time.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
