CIC supports deploying multiple CIC Agents in a High Availability (HA) configuration to help ensure continuity of services even when you experience hardware glitches.
CIC Agents are installed when you create an Access Point. (See Managing Access Points.) When you install more than one CIC Agent, you are implementing High Availability Agents (HA Agents). Up to 5 HA Agents (called Nodes) can service a single Access Point. Each of these Nodes can provide connectivity between the File System or Database Endpoint and CIC. If the currently active Node goes down (which means the Node is no longer sending a "heartbeat" to CIC), CIC routes requests to other Nodes in the Access Point, providing continuity of service. A Node is considered "up" again when CIC receives a "heartbeat".
CIC selects the first CIC Agent installed to be the leader node. If the leader node becomes unavailable, a new leader is automatically selected by CIC. The leader node handles Jobs started using the Run Now button (Data Flow page), Jobs that run on a Schedule, Remote Browse connections, and DB Proxy connections.
The "Send Logs to Cleo" functionality includes the logs from all nodes.
Note: HA Agents do NOT provide mid-Job switchover or increased throughput.
High Availability Agent Requirements
HA Agents have the following requirements:
- All Nodes for a given Access Point must be installed on separate servers that use the same operating system (Windows or Linux).
- The infrastructure representing the endpoints, databases, and file systems, MUST be shared with all Agents.
- File system paths must be the same for all Agents.
- Database TCP ports and addresses must be reachable by all Agents in the cluster
Upgrading to HA
Access Points are automatically upgraded to support HA when their agents are updated to an HA-capable version, which starts with v1.1 of the CIC Agent. See Installing, Updating, and Uninstalling CIC Agents.
Note: When performing updates to an Access Point, only one Agent is in an "update" state at a time.