Skip to main content

Troubleshooting

This page maps the symptoms you see when running Clustron Zaris to their likely causes and the fixes that resolve them. It is organized by where the problem shows up: connecting to managers, creating and starting stores, forming the cluster, connecting clients, and reading metrics. Start with the table that matches your symptom.

Before diving in, two checks resolve a large share of problems: confirm you have an active workspace in the current session (Get-ZrWorkspace), and confirm the relevant ports are open. Use Get-ZrStore at any point to inspect the current state of the cluster.

Cannot connect to a manager

A cmdlet reports no manager connection, or a workspace shows fewer reachable managers than you listed.

SymptomCauseFix
Timeout or connection refusedThe Management Service is not runningConfirm Clustron is installed and the service is started on the target server
Refused on port 7801Firewall blocks the management portAllow inbound TCP 7801 on the server and any firewall in between
Some managers unreachable in the rosterAn endpoint is wrong or unreachableVerify each address and port; check connectivity between machines
Works locally, fails from another hostYou used localhost or 127.0.0.1Use the server's real IP or hostname for multi-server clusters

Store fails to create or start

New-ZrStore reports an error, or Start-ZrStore leaves a node down.

SymptomCauseFix
"No active Zaris manager connection"No workspace is activeActivate a workspace first, or pass -Managers
"cannot be created here — this is an attach-mode manager"The manager runs in attach modeAdopt the externally-run store with Register-ZrStore
Error on one manager, success on othersThat manager is unreachableConfirm every manager is running and reachable, then retry
Node fails to launch on startA cluster or client port is already in usePick free -BaseClusterPort / -BaseClientPort ranges; avoid overlapping stores
"port range … overlap" on createThe base ports are too close for the per-server process countWiden the gap between -BaseClusterPort and -BaseClientPort

Nodes do not join the cluster

The store starts but Get-ZrStore reports Partial, or only some nodes appear.

SymptomCauseFix
Only some nodes are visibleCluster ports blocked between machinesOpen the cluster range (default 78117850) across the network
Cluster never formsNot all managers were in the workspace when the store was createdActivate a workspace with every manager, then re-create the store
Nodes on other servers unreachableWrong or unreachable addressesUse reachable IPs or hostnames — not localhost — and verify connectivity

Clients cannot connect

An application or the ClientShell fails to reach the store.

SymptomCauseFix
Connection failsThe store is not runningStart the store with Start-ZrStore; confirm Running with Get-ZrStore
Connection refused on the client portClient ports blockedOpen the client range (default 78617899)
Wrong endpointIncorrect host or client port suppliedConnect with the correct <host>:<clientPort> for a running node

No metrics displayed

Watch-ZrStoreMetrics shows no rows, all-zero totals, or stops updating.

SymptomCauseFix
No rows or all-zero totalsThe store is stopped, or no client is sending trafficConfirm Running with Get-ZrStore; generate load with client operations
Totals rise but one node column stays flatThat node is down or unreachableCheck the node with Get-ZrStore; start it if stopped
View stops updatingClient activity stopped, or a manager became unreachableVerify clients are active and the managers are reachable
"No managers connected"No workspace is activeActivate a workspace, or pass -Managers

General guidance

A few habits prevent most of the problems above:

  • Always activate a workspace before any other admin command, and confirm its managers with Get-ZrWorkspace.
  • Verify port and firewall settings early — especially the cluster and client ranges across the network on multi-server clusters.
  • Use Get-ZrStore to inspect cluster state whenever behavior is unexpected.
  • Start with a single-server cluster to validate your workflow before scaling out to multiple servers.

Next steps