Skip to main content
Start with the resource’s status, recent activity and logs. For application-specific configuration, also consult its marketplace page and upstream documentation.

​
VMs

Open the VM’s activity and error details to see the last completed step. Provisioning includes several background operations, and duration can vary.If progress stops or an error persists, contact dashboard Support with the instance ID and recent activity. Preserve any existing data before deleting or rebuilding a VM; retrying by deletion is not a general recovery step.
Check the organization’s credit, spending controls and the instance’s current status. An operation already in progress can prevent another start request. At zero credit, a VM may have been terminated rather than merely stopped; see When your credit runs low.Follow the error shown by the dashboard. If it persists after the prerequisite is resolved, contact support with the instance ID.
Review CPU, memory and disk metrics alongside application logs. Sustained resource pressure can indicate a sizing issue, but can also come from an application error, workload change or full disk.Check for crash loops and unexpected processes before deciding to resize. A resize requires an interruption; it is not a guaranteed fix for application problems.
  1. Use the VM’s current public IP and the private key matching a public key supplied before VM creation.
  2. For the standard provisioned root account, use ssh root@<public-ip>.
  3. Check that the VM is running and that firewall rules allow your current source IP to reach SSH.
  4. Check the SSH service and any guest-level firewall or access changes you made.
The dashboard Terminal can provide another access path when the VM and guest agent are reachable. See Terminal access.

​
Apps

Open the install’s logs and error details. Image downloads, image access, configuration and container startup can each fail independently.
  • Image denied / unauthorized: check the image name, version and registry access requirements.
  • Port in use: identify the conflicting listener before changing ports or restarting services.
  • Permission denied / EACCES: check the affected path, mount and ownership against the app’s requirements. Do not broadly relax permissions to hide the error.
Resolve the reported cause before retrying. Restart and upgrade are different operations; an upgrade can change the application version or configuration. Preserve important data before disruptive changes.
Check the current URL shown for the install, the VM and container status, and the app’s listening port. Read startup logs for crashes or failed health checks. A routing, tunnel or application problem can all prevent the URL from working.If the container is healthy but the URL still fails, record the URL, time and response details for support. See the app’s guide, such as OpenClaw troubleshooting, for its own network settings.
  1. Confirm the sender uses the current app URL, endpoint path and required authentication.
  2. Send a non-destructive test event and inspect the sender’s delivery status and response.
  3. Compare that result with the app’s logs and webhook configuration.
An absent log entry does not prove a DNS failure: some apps do not log every request. A 4xx or 5xx response can come from the app or an intermediary; use the response and logs to locate the failure.

​
Billing & credits

Check the instance activity, snapshots and organization balance. Zero-balance handling can terminate VMs and attempts to retain a system-disk snapshot for seven days. Snapshot creation can fail, so recovery is not guaranteed.Top up and check for an available, ready snapshot before its deadline. A termination snapshot’s deadline is not extended by topping up. See When your credit runs low.
Review the recent attempt and error under Settings → Payment methods. Possible causes include a declined or missing card, a payment needing action, or the monthly recharge cap.Correct the relevant setting and consider a one-time top-up while resolving it. Check suspended VMs separately before using start or resume. Do not rely on an automatic retry or a fixed grace period to protect the environment.
Confirm that the payment completed and that you are viewing the organization you topped up. Refresh the balance and payment history. If a completed payment remains missing, contact support with the payment reference, amount and time. Do not send card numbers or security codes.
Check the payment-history entry for an available invoice link. Not every payment path or older payment has one. Update Settings → Billing details before future payments and contact support to review a missing or incorrect document.

​
Account

Open Forgot password and enter your account email. Use the latest reset email; reset links expire after one hour. If a link has expired, request another.
Check spam and confirm the registered email address. Use the verification resend control in the dashboard. Requests are rate limited; if a resend is rejected, wait before trying again. Contact support if you still cannot verify the account.
Confirm that your authenticator’s clock is synchronized and that you selected the correct Suji entry. If available, use one of your unused backup codes at sign-in.Contact support if you cannot recover access. Do not send your authenticator secret or backup codes in the support request.
An authorized owner or admin can review organization membership and invite the colleague with the appropriate role. Check their project access as well. Review a departing member’s responsibilities before removing access; do not share their login credentials.

​
Network / firewall

Check the VM’s current address, the firewall direction, protocol, port and allowed source IP. Then check the guest firewall and the application’s listener. For TCP, ss -tlnp from the VM terminal lists listening sockets.For a direct public port, verify the container’s port publication and intended bind address. Tunnel-routed apps use a different path; do not expose a private listener publicly just to troubleshoot it. See Firewall.
Check the application’s SMTP host, port, TLS mode and authentication, then review its error and network rules. If the required SMTP port is blocked, use a supported submission port or an HTTPS email API offered by your mail service. See Firewall for network restrictions.

​
Contact support

Open dashboard Support with the organization, relevant resource IDs, time of the problem, steps to reproduce and a short excerpt of relevant logs. Redact passwords, tokens, personal data and other secrets from logs and screenshots.