35t DAQ
Revision as of 14:50, 21 October 2014 by MaximPotekhin (talk | contribs)
Recent Meetings
- Joint DAQ/S&C/Offline meeting on October 20, 2014. Information includes a long list of questions that were brought up and clarified.
Recent documents
- DocDB 9677, Proposed Run Modes for the 35t test
- DocDB 9785, Report from RCE-Backend Mini-workshop
- DocDB 9591, 35t Phase-II Testing for August Readiness Review
- DocDB 9823, summary for plans in mid-Fall 2014
Run Modes (per DocDB 9677)
| No. | Mode | Description | Comment |
|---|---|---|---|
| 1 | Continuous main DAQ mode | Similar to final 'triggerless' far detector running mode, i.e. parallel trigger farm looks for nice muons in real time. Needs ZS of TPC data to work. | --- |
| 2 | Triggered main DAQ mode | Main selection is external trigger counter | --- |
| 3 | Immediate triggered mode | Use TOC triggered mode (only outputs data in window) or SSP triggered mode to avoid bottleneck. | This is a fallback option of there is bottleneck in either RCE or SSP data. |
| 4 | Triggered window mode | TPC in triggered non-zero suppressed mode | Use before we are happy with ZS in RC |
| 5 | Wide-window burst mode | TPC in burst mode with the longest window possible (512MB buffer = 0.5s). | Special run mode for offline ZS studies |
| 6 | Burst mode | Collect data in certain time window (could either determine this from a Nova time hash function, or using Penn board trigger input | Fallback before ZS works in RCE, this is also the mode for measuring noise, i.e. this is the mode for November vertical slice test |
Issues
- Purposes and architecture of the proposed online farm
- Design of the metadata
- Handling TPC stream data within the framework which is trigger-oriented, issues of timing and multiple interaction in temporally close slices
- Possible necessity to duplicate parts of raw data to create overlaps, in order to make practical operation of the online farm
