After you create a Provider High Availability (HA) cluster, VMware Data Services Manager provides status of all the Provider nodes in the cluster. You can view the status of all the Provider nodes, analyse the status, and take actions if required. Easy access to the status of the Provider nodes helps you to manage the cluster better and troubleshoot, if required.
Status of Provider Nodes
The following table lists different status of Provider nodes in an HA cluster:
| Status | Description | Implication | Recommendation |
|---|---|---|---|
| Active Mode | Node is healthy and all services are up and running. | Only an Active node can consume updates. | Apply VMware Data Services Manager updates for provider nodes only when they are in Active mode. At least one Secondary node of the cluster must be in Active mode to ensure smooth functioning of the cluster. |
| Out of Sync | Only a Secondary node can have this status. Node does not have the latest cluster information and is not able to catch up with the Primary node. | Node is eligible for deletion. | Delete the node before performing any operation, for the smooth functioning of the cluster. |
| Critical | Apart from the application management service (apms) and vpostgresql service, any other VMware Data Services Manager (provider, update manager, and so on) or system (coredns, rabbitmq, and so on) service is down. | Node is unable to consume any updates. | Execute the Provider Appliance Health Status API (/appliance/status) to understand the details, and then take action as required. |
| Lost Connectivity | Node is unable to connect to the Provider appliance because apms or vpostgresql service is down. | Node is down, unreachable, or unable consume any updates. | Before you update or edit the cluster, troubleshoot to check if the Provider node is up, if apms is up, and if vpostgresql service is up. Restart the service that is down. |
| Maintenance Mode | Node is consuming VMware Data Services Manager updates. | Wait until VMware Data Services Manager updates are completed. | Do not perform any Add or Delete operations on the Provider Node because it may lead to instability of the cluster. |
| HA Config in progress | Provider HA configuration is in progress. | Node cannot consume any updates. | Wait until the Provider HA configuration is complete. Do not attempt any operation on the node. |
| Unknown | VMware Data Services Manager is not able to provide any status of the node because root password of the Provider appliance has expired. | Node cannot consume any updates. | Change the root password immediately for the smooth functioning of the cluster. |
Note:
During Add, Delete, and Promote operations in a Provider HA cluster, VMware Data Services Manager attempts to reconfigure all the nodes of the cluster. If there is failure to reconfigure any node, the status of that node becomes Out of Sync.
Configure the Provider Load Balancer to forward API requests only to the nodes that are in Active mode. You can use the /appliance/provider-ha-nodes API or /appliance/status API to determine the health of a Provider node.
If all the Secondary nodes of a cluster are down or have Lost Connectivity as the status, Add, Delete, and Promote operations in the cluster fail. Before performing these operations, troubleshoot to bring up the secondary nodes or remove them from the cluster.
