Avoid monitoring group update failure for single register access error
Description
In the current implementation, a single failure during the monitoring group updates prevents the overall group update. While technically not problematic (the last update timestamp is recorded), that behavior can create problems (e.g. the absence of valid FPGA Core Temperature for the whole AMCMonitor
).
It looks like the heuristic currently in place (derive the OptoHybrid mask from the FW_VERSION
readout validity) isn't strong enough to cover all cases during the real operations with collisions.
Possible fixes
- Find a better way to update/derive the OptoHybrid mask
- Allow the group update not to fail for a single register access error