Troubleshooting
Dashboard Not Accessible​
Symptom: Browser cannot reach https://<head-node-ip>:30082
Check:
- The head node is powered on and the network cable is connected
- You are on the same network as the management interface, or have a routed path to it
- No firewall is blocking ports
30082(HTTPS),30080(HTTP), or30083(REST API) between your workstation and the head node
If the appliance was just powered on, wait 2–3 minutes for the management software to start.
System State Is Degraded or Failed​
A banner at the top of every page lists each cause with a suggested remedy. Each entry may offer one or more recovery actions, ordered least disruptive first. Before running one, the dialog states what it does, whether data access is interrupted, how long it takes, and whether it is reversible; disruptive actions require an explicit acknowledgement.
Some conditions have no safe action from the Dashboard. Those are presented as a host procedure with a route to System → Support instead of a run button.
A Node Shows Non-Ready Status​
Navigate to Hardware → Nodes and click the affected node.
| Status | Likely cause | Action |
|---|---|---|
| Initializing | Node is still starting up | Wait 5–10 minutes after power-on |
| Degraded | Hardware issue, or storage temporarily unreachable on the fabric | Check the node's drive and link status; unreachable storage is recovered automatically |
| Failed | Node cannot be reached | Verify network cable and power; check the node's status light |
Nodes also report a reachability status — Synced, Recovering, or Unreachable.
Recovering is normal and self-clears; a node that stays Unreachable needs the fabric link
checked.
A Storage Engine Is Not Responding​
When an engine that handles drive I/O stops responding and automatic recovery has given up, a warning banner appears for all users. Administrators see a Recover engine button on the banner, which restarts that engine after a confirmation step. I/O to the affected drives may be briefly interrupted.
Storage Capacity Is Lower Than Expected​
Navigate to Hardware → Drives.
- Every data drive should show Healthy and In Cluster
- A drive showing Failed can be replaced from this page — see Replacing a Drive
- A drive marked System Disk is the boot drive and is never cluster storage
If drives are healthy but capacity is still short, check Storage capacity is lower than expected.
Volume Group or Volume Stuck in Provisioning​
Provisioning normally completes within a few minutes. Operations that stall on the storage backend now recover on their own within minutes, and a volume that still cannot finish is marked Failed with a message rather than staying stuck.
If an item stays in Provisioning for more than 10 minutes:
- Navigate to Storage → Volume Groups (or Volumes) and click the stuck item
- Read the status message
- Verify the target nodes are Ready under Hardware → Nodes
Common causes:
- Insufficient free capacity — check used vs. total capacity
- A node went offline during provisioning
Locked out of the Dashboard​
If you still have Dashboard access, change your own password from My Profile → Change Password, or have another administrator reset the account from Identity & Access → Users & Groups.
If no administrator can sign in, an operator with shell access to the appliance can restore the administrator account:
sudo mass-ctl reset-admin-password --restore-access
This re-activates the administrator identity and restarts the management service. Contact MangoBoost support if you do not have shell access to the appliance.
Dashboard Shows Stale Data​
Most views refresh on their own. If data still looks stale, reload the page; if that does not help, log out and back in.