Troubleshoot by symptom
Start with the observed symptom. Keep the current state and error before taking another action.
Connection issues
Section titled “Connection issues”| Symptom | Where to check | Next step |
|---|---|---|
| SSH timeout | Your computer: endpoint and IP family; Panel: running state | Follow SSH diagnostics. |
| SSH refused | Guest console: SSH service and port | Inspect the OS-specific service; do not open every firewall port. |
| Permission denied (publickey) | Your computer: selected key; guest: authorized account | Match the existing key and username before generating or replacing keys. |
| RDP unavailable | Guest console: Windows state and approved endpoint | Follow RDP diagnostics. |
| No console | Panel permission and console connection state | Follow console help. |
| No assigned IP | Server network information | Ask support; do not choose an arbitrary address from the gateway subnet. |
| NAT connection fails | External address and external port | NAT metadata alone does not create gateway rules. See network. |
Do not disable host-key checks, enable root password login, or disable all firewall rules as generic remedies.
Power, performance, and missing observations
Section titled “Power, performance, and missing observations”| Symptom | Safe next action |
|---|---|
| Node offline or state unknown | Record the time and contact support; avoid repeated power requests. |
| Server remains busy | Check Activity and wait for the current job; report a persistent failure. |
| No metrics | Check observation freshness and selected range. Missing data is not zero usage. |
| Slow application | Compare resource observations with guest-side evidence; do not assume a plan upgrade fixes it. |
| Installation failed | Record the progress stage and error; ask support before retrying or ordering again. |
See Telemetry and Power. Any guest diagnostic command runs inside the guest, not on the provider host. Use a tool already supported by your operating system; do not install a privileged collector for a simple report.
Recovery and access
Section titled “Recovery and access”| Symptom | Safe next action |
|---|---|
| Backup pending | Check busy state and quota; avoid a second overlapping request. |
| Backup failed or incomplete | Preserve the error; do not restore an unverified archive. See backups. |
| Snapshot unavailable | Confirm entitlement and storage support; a snapshot is not an external backup. |
| Invitation expired | Ask the inviter for a fresh authorized invitation. Do not share the old link. |
| API token expired or rejected | Check expiry, scope, and account access; rotate privately. |
| Missing server or button | Check workspace, effective permissions, and hosted feature availability. |
| Paid order still pending | Check the invoice and contact support with its private reference; do not pay twice. |
Guest software tutorials
Section titled “Guest software tutorials”For Docker-specific IPv6 setup, read the existing Docker tutorial. Tutorials describe guest software and do not guarantee that a particular image, route, or entitlement exists on your service.
If the failure boundary is the host, network assignment, console broker, payment integration, or entitlement, contact support. Send a redacted error and time zone, not credentials or full browser captures.