Diagnosing and fixing problems
Viewing cluster health using "diagnostic" command
The diagnostic command displays the online, activation, health, restart, state and configuration records for each node in the cluster.
Syntax:
diagnostic -connect-to <hostname[:port]>
Parameters:
-connect-to <hostname[:port]>The node to connect to
Examples:
The following example shows the execution of the diagnostic command on a single-stripe cluster containing two nodes, with one node activated and online, and the other node offline:
Diagnostic result:
[Cluster]
- Nodes: 2 (node-1Q@10.0.0.1:9410, node-2@10.0.0.2:9510)
- Nodes online: 1 (node-1@10.0.0.1:9410)
- Nodes online, configured and activated: 1 (node-1@10.0.0.1:9410)
- Nodes online, configured and in repair: 0
- Nodes online, new and being configured: 0
- Nodes pending restart: 0
- Configuration state: The cluster configuration is healthy. New configuration changes are possible.
- Configuration checkpoint found across all online configured nodes (activated or in repair): YES (Version: 1, UUID: e9e0b92c-db10-4275-842e-a8bf6210e97b, At: 2020-09-03T13:06:07.935, Details: Activating cluster: my-cluster)
[node-1@10.0.0.1:9410]
- Node state: ACTIVE
- Node online, configured and activated: YES
- Node online, configured and in repair: NO
- Node online, new and being configured: NO
- Node restart required: NO
- Node configuration change in progress: NO
- Node can accept new changes: YES
- Node current configuration version: 1
- Node highest configuration version: 1
- Node last configuration change UUID: e9e0b92c-db10-4275-842e-a8bf6210e97b
- Node last configuration state: COMMITTED
- Node last configuration created at: 2020-09-03T13:06:07.935
- Node last configuration created from: admin-machine
- Node last configuration created by: admin
- Node last configuration change details: Activating cluster: my-cluster
- Node last mutation at: 2020-09-03T13:06:08.035
- Node last mutation from: admin-machine
- Node last mutation by: admin
[node-2@10.0.0.2:9510]
- Node state: UNREACHABLE
- Node online, configured and activated: NO
- Node online, configured and in repair: NO
- Node online, new and being configured: NO
Fixing cluster health using "repair" command
The repair command can be used to fix cluster configuration inconsistency problems. The result of running this command on a healthy cluster is a no-op.
Important:
Please contact Software AG support before running this command
Syntax:
repair -connect-to <hostname[:port]> [-force commit|rollback|reset]
Parameters:
-connect-to <hostname[:port]> The node to connect to
-force commit|rollback|reset Force a commit, rollback or reset operation