Common Questions
Find the question that matches what you see. Record the exact error and the time it occurred before restarting anything.
Why does Management not start?
-
Open the Management VM’s hypervisor console and select Refresh Status.
-
Check that the intended ISO remains attached and the state disk is present. If the console reports missing, duplicate, corrupt, or incompatible state, preserve the disks and read Show Recovery Status. Do not initialize a disk to clear an unexplained state error.
-
Open Show System Information. Check the active interface, address, gateway, DNS, and synchronization with the configured time source. Management startup requires both active network configuration and synchronized time.
-
Inspect Logs > View recent MTG host service logs and View warnings and errors.
-
Correct the identified host problem. Use Retry State Discovery for a corrected state-discovery problem, or Start Management Application when the host is ready and Management is stopped.
The host retries temporary container-inspection timeouts within the startup deadline. Follow Refresh Status and the host logs while startup continues; do not restart solely because one inspection timed out. A stopped container or a non-timeout inspection failure still requires investigation.
Verify that the console reports Management ready and that its displayed HTTPS URL opens. If startup fails again, collect the host evidence described in diagnostics and contact support. Decide whether to restart a running but unhealthy container after checking active operations. Do not interrupt an operation just because a health check failed.
Why can’t I open Management in the browser?
Compare the browser URL with the console’s public Management URL, including its configured HTTPS port. Check that the hostname resolves to the Management VM, and that the browser can reach that address through the required firewall rules.
For a certificate error, check hostname coverage, trust, expiry, and system time. Compare the certificate with the configured Management TLS mode. Use Manage Certificates for a certificate correction and verify the new certificate afterward. For ACME, the certificate authority must reach Management on TCP 80.
Changing the host address does not change the fixed public Management hostname. Correct DNS and network routing to preserve that identity. See Management setup.
Why does Management require administrator access?
Confirm that the signed-in user is a Management administrator. Management takes this status from its Gitea-backed session; an ERS or Grafana account is not sufficient. Sign in with the appropriate account and reopen the task.
For lost administrator access, use Management reconstruction. The Management host console has no browser-account password reset workflow.
Does the appliance stop when Management is disconnected?
No. The appliance continues running its workloads and scheduled backups. Management needs active cluster access for administrative work. Check appliance health separately from Management connection status. See Reconnect an installed appliance and Inspect health, logs and audits.
Why has cluster access expired or an administrator tool become unavailable?
Management holds uploaded credentials for a six-hour access period. Browser sign-in and ordinary use do not extend it. Reconnect after expiry or a Management restart, or renew access before it expires using the same appliance’s bundle.
Use Reconnect with the retained artifact bundle. Check that Management can reach the appliance and that the bundle belongs to the intended cluster. Then reopen the failed task.
If Management reports Access removed from Management; firewall closure unconfirmed, upload the saved bundle to reconnect and reconcile access. If renewal is unavailable, wait for the running automated operation to finish. Gitea availability and cluster health are separate checks.
For Settings unavailable after reconnecting, use Open Home to check whether Git configuration needs recovery. Restore the appliance’s published Git history when offered, then reopen Settings. Do not overwrite configuration with guessed defaults.
Why can’t I select a release or recovery bundle?
Open Settings > Bundles and check its type, availability state, and recorded release. A fresh installation needs a release bundle; an update needs an exact source-to-target update bundle. Recovery needs a stored full backup whose release matches the selected release bundle, plus matching artifacts.
If upload or retrieval is blocked for lack of Management space, use storage maintenance. If a registry retrieval fails, check the configured endpoint, trust, and credentials. Local upload remains available when the required release media has been supplied offline.
Why does node discovery or disk selection fail?
Check that the intended replacement or new appliance machines are powered on with the selected boot media and reachable network configuration. Compare discovered node identity and addresses with the deployment plan.
If an expected disk is absent or its identity differs, check the VM’s attached disks and controller configuration before proceeding. Return to discovery and select the actual intended disks. Never choose another visible disk merely to satisfy the wizard.
Record the failed stage and use diagnostics. Cancellation stops execution; it does not restore disks already changed. See the installation procedure for checks and retry instructions.
Why hasn’t my configuration change taken effect?
A save records the desired configuration. Management publishes it and the appliance applies it through GitOps. Verify the affected service after a change.
-
Check that the appliance is connected and no update is holding configuration publication.
-
Inspect any dashboard publication failure notice and follow its Gitea commit link.
-
Correct invalid configuration through the appropriate task or a new configuration commit.
-
Inspect Grafana and workload logs for runtime health and reconciliation failures.
Saving or publishing configuration does not confirm application health. A successful later publication clears a publication failure notice. See GitOps ownership.
Why is my backup missing or its download failing?
Select Refresh backups and check both full and database filters. Only eligible completed backups appear. Review capture and storage health if no completed backup is available.
If the selected file expired, choose another available backup. If access expired, reconnect and restart the download. Check completion in the browser download manager. Check external destination files separately from the local list. See backups and retention.
Is a full backup enough to recover the appliance?
Recovery also requires the matching artifact bundle, matching release bundle and compatible Management. The artifact bundle contains private recovery keys. Retain these inputs outside the appliance. See Recover the appliance.
Are metrics, logs and external HSM keys included in backups?
Full appliance backups exclude metrics history, log history and Grafana’s local database and UI-created state. External HSM keys require the HSM vendor’s backup procedure. See What a backup protects and Customer monitoring and Syslog.
Should I recover the appliance if Management was lost?
If the appliance survives, reconstruct Management and recover its configuration from the appliance. Full appliance recovery is for lost appliance machines or data. Use Choose a recovery path to select the appropriate procedure.
Why is my diagnostics archive incomplete, or why did collection time out?
Download the available partial archive before starting another collection. Missing target-cluster access, an unreachable component, or collection limits can produce a partial result. Restore connectivity and reconnect before collecting again if target evidence is needed.
A complete diagnostics archive means the applicable evidence was collected; it does not mean the system is healthy. If no archive is available, preserve the displayed error and collect host logs. See diagnostics.
What should I collect before contacting support?
Record the exact error, its time, the affected operation and the displayed appliance state. Preserve available diagnostics and operation logs before retrying or restarting. Follow Collect diagnostics and Contact MTG support.