OpcUa Cluster Item
An item used to configure a cluster of two or more redundant OpcUa servers. This item defines the cluster and its failover behavior, while the connection information for each server in the cluster are defined by OpcUa connection items.
Each OpcUa cluster can be selected as a server configuration option in the OpcUa and OpcUaProxy modules, thereby giving these module types transparent failover support.
A failover between servers can occur when the ServiceLevel reported by each server changes. The ServiceLevel is a number from 0 to 255 with the following semantics:
- Levels 0-1: The server is not able to provide any data from its data sources
- Levels 2-199: The server has lost connection to some, but not all, of its data sources
- Levels 200-255: The server is connected to all of its data sources
There are different failover rules for each ServiceLevel group:
- The active server has ServiceLevel 0-1: a failover occurs if another server has a ServiceLevel greater than 1.
- The active server has ServiceLevel 2-199: a failover occurs if another server has a ServiceLevel greater than 199, or if another server has a ServiceLevel greater than the ServiceLevel of the active server plus the value of the 'Failover deadband' property.
- The active server has ServiceLevel 200-255: failovers do not occur
Additionally, failovers never occur more frequently than specified by the 'Failover interval' property.
The OpcUa cluster item has the following standard properties:
| Name | Description | ID | Flags |
|---|---|---|---|
| Name | The name selected for the cluster. This name will become available as a Server configuration option in the OpcUa and OpcUaProxy modules. | ||
| Value | The server endpoint currently having the highest ServiceLevel for this cluster (read-only). The active OpcUa connection of the cluster does not necessarily change even when this value changes. | 2 |
The OpcUa cluster item has the following advanced properties:
| Name | Description | ID | Flags |
|---|---|---|---|
| Failover timeout | Number of seconds to wait for additional servers to come online (after the first server reponse), before selecting a server as the current endpoint for the cluster. | 10100 | Persisted |
| Failover interval | Minimum number of seconds between failovers. | 10101 | Persisted |
| Failover deadband | Minimum ServiceLevel difference required to trigger a failover (only applies to servicelevels below 200). | 10102 | Persisted |