How does OPC Expert handle asynchronous writes and data changes during Warm or Cold redundancy when the standby has no active data connection?
In Warm and Cold redundancy modes, the standby OPC server is not actively subscribed to the underlying control system until a failover is triggered. During this period, write operations and data changes directed at the standby server are not forwarded to the PLC or DCS because no active data channel exists between the standby OPC server and the hardware. Writes made through OPC Expert are directed to whichever path is currently active, which is the primary under normal operating conditions.
How Write Operations Are Routed
OPC Expert acts as an intermediary between your application and the OPC servers. When your application issues a write command through OPC Expert’s embedded OPC server output, that command is directed to the currently active data source. In Warm and Cold modes, the active source is the primary server. The standby server, because it has no live subscription to the control system, cannot accept or relay write commands to the hardware during normal operation.
This is an important architectural consideration for applications that write setpoints or control outputs through OPC. If write continuity during a failover is critical, the Hot redundancy mode is the more appropriate choice. In Hot mode, both servers maintain active connections to the hardware, so the standby server can accept and relay write operations immediately after a switchover without a gap.
Behavior During a Warm Mode Transition
When a Warm mode failover occurs, OPC Expert signals the standby server to begin active data collection. During the period between failure detection and the moment the standby server establishes its subscription cycle with the control system, write operations cannot be confirmed as delivered. The length of this window depends on how quickly the standby OPC server can initiate its connection to the underlying hardware. In practice this is typically a matter of seconds, but it is not zero.
OPC Expert surfaces this transition state through its diagnostic output, reporting the current server action and status for both the primary and standby paths. This gives operations staff visibility into exactly what is happening during a switchover rather than leaving the system state ambiguous.
Behavior During a Cold Mode Transition
In Cold mode, the standby server is not running at all when the primary is active. After a primary failure, OPC Expert waits for the standby server to start before it can establish a connection and begin data collection. Any write operations issued during the period between the primary failure and the standby becoming available cannot be relayed to the control system. The duration of this window is determined primarily by how long the standby server takes to start and register.
Cold mode is generally used in architectures where the control system vendor does not permit more than one OPC server to run simultaneously. In those cases, the write continuity gap is an accepted constraint of the architecture rather than a limitation of OPC Expert itself.
Related Tools
OPC Redundancy is the OPC Expert feature that manages the connection state between OPC Expert and the primary and standby servers, including routing read subscriptions and write commands to the currently active path.
OPC Bridge is a separate OPC Expert feature designed specifically for transferring data between OPC servers. It is relevant when the goal is data replication between servers rather than client-side failover, and includes its own fault detection and redundancy mechanisms.
Conclusion
In Warm and Cold redundancy modes, writes are directed exclusively to the active primary server path. The standby server does not relay write commands to the control system until a failover occurs and the standby establishes an active data connection. Applications with strict write continuity requirements should use Hot redundancy mode, where the standby maintains a live connection to the hardware at all times and can accept writes immediately after a switchover.