Some ideas on a possible new server architecture for a world without RF flat-tops:
- Can we mark each entry in the 9 MHz timing table with one bit as "good beam required here"? Or do we establish a separate software protocol for that?
Motivation and use cases
Necessary software modifications
|Bunch pattern server||Develop & implement an additional mechanism to mark subpatterns as "good beam" regions||~5 days|
|HLC machine library||Adapt EnergyGainInfo interface|
|HLC tracking library||Adapt ComponentList::set_momentum_from_energy_gain_info(), StandingWaveCavity||~1 day|
The SumVoltage Server is not anymore the master of its settings, but will receive an energy-table from the HighLevel Configurator.
The LinacEnergy Mgr is not anymore the master of its settings, but will receive an energy-table from the HighLevel Configurator.
Assuming, the LLRFctrl server still allows to use one SetPoint value to recover a RF-Station, there are not much changes needed. Obviously, the FSM will not keep anymore the deprecated information about FlatTop 2 and 3. This Beam-Area table should be kept somewhere else. At HighLevel Configurator or the LLRFctrl-server. Which means, the FSM recovers a RF-Station partially only.
In case, the FSM needs to keep this new Beam-Area table, the effort will be several weeks including changes of the test environment at the VXFEL
|Install/Roll-back||A Install/roll-back mechanism is required for DOOCS server, DOOCS server configuration, firmware and jDDD panels|